"RNase H2 Complex and Genes Therefor"

The present invention describes genes encoding components of Ribonuclease (RNase) H2 , recombinant RNase H2 and an assay to identify agonists, antagonists and modulators useful in manipulating the immune response, for example to treat viral infection and/or autoimmune disease.

RNase H is an enzymic complex responsible for endonucleolytic cleavage of ribonucleotides from RNA/DNA complexes. RNase H is frequently used in molecular biology to degrade the RNA template subsequent to reverse transcription of cDNA.

Two classes of RNase H exist (types 1/1 and 2/11) with distinct biochemical properties. RNase H2 is the major source of cellular RNase H activity in both humans and yeast. RNase H2 is proposed to function in the removal of lagging strand Okazaki fragment RNA primers during DNA replication, and for excision of single ribonucleotides in DNA-DNA duplexes. In S. cerevisiae Rnh2Bp and Rnh2Cp copurify with the catalytic subunit Rnh2Ap and together are sufficient to reconstitute RNase H2 activity. In humans, the RNase H2 catalytic subunit, RNASEH2A, has been identified by biochemical purification and has clear sequence homology with its yeast ortholog. However, no orthologs outside yeast have been identified for the Rnh2Bp and Rnh2Cp subunits, though a second human protein AYP1 (now termed RNASEH2C) was copurified with the RNASEH2A protein (Frank et al., PNAS USA 95:12872-7, 1998).

Aicardi-Goutieres syndrome (AGS) is an autosomal recessive genetic disorder of unknown aetiology. AGS presents clinically as severe neurological dysfunction, progressive microcephaly, spasticity, dystonic

posturing, and psychomotor retardation. Death frequently occurs in childhood (Goutieres, Brain Dev 27:201-206, 2005). AGS demonstrates strong phenotypic similarities to congenital viral infections of the brain (Aicardi & Goutieres, Ann Neurol 15:49-54, 1984), with the raised IFN-α levels noted in AGS also similar to host immune response following viral infection. Consequently, there remains much debate in the literature regarding the aetiology of AGS, and in particular whether AGS is a consequence of genetic susceptibility to viral pathogens or of aberrant regulation of host immune response.

Crow et al., (Am J Hum Genet 67:213-221 , 2000 and J Med Genet 40:183-187, 2003) identified two AGS gene loci, namely AGS1 on chromosome 3p21 and AGS2 on chromosome 13q14.3 (AIi et al., J Med Genet 43(5):444-50, May 2006. Epub 20 ^th May 2005).

The inventors have now found that AGS2 is a component of the human RNase H2 complex and have fully identified this gene. AGS2 had a ubiquitous expression pattern, no predicted domain-structure, or human paralogs from which to infer function. A distant ortholog in S. cerevisiae (<5% amino acid sequence identify) was located. The inventors have also identified the AGS3 and AGS4 genes. The expressed proteins of these three AGS genes together form part of the RNase H2 complex.

AGS has many phenotypic similarities to viral infection and this resemblance is not confined to the brain with some cases having extra- neurological features (such as thrombocytopenia, hepatosplenomegaly and elevated hepatic transaminases), suggesting overlap with other genetic disorders that also mimic viral infection.

Similarities between AGS and Systemic Lupus Erythematosis (SLE) have also been noted (see Alarcan-Riquelme, Nat Genet 38:866-867, 27 ^th July 2006 ). In particular acral vasculitic skin lesions occur in SLE which exhibit immunoglobulin deposition at the dermal-epidermal junction (see Crow et al., Nat Genet 38:917-920, 2006). A recently reported familial cutaneous lupus erythermatosis disorder with identical skin lesions colocalises with the AGS1 gene (see Lee-Kirsch et al., Am J Hum Genet 79:731-737, 19 July 2006). Additionally, hypergammaglobinema and autoantibodies have been reported in AGS (see Crow et al., Nat Genet 38:917-920, 2006. Furthermore, raised levels of CSF Interferon alpha are found in both cerebral lupus and systemically in SLE. Intracranial basal ganglia calcification also occurs in up to 30% of patients with cerebral SLE.

As a consequence of the observations by the inventors, it now appears for the first time that AGS, SLE and viral infections have a common underlying aetiology involving RNase H2.

Importantly, the inventors have identified in the AGS4 gene, the G37S mutation in family F39, in which affected individuals exhibited features of both pseudo-TORCH and AGS, and which indicates that these disorders share a common molecular basis. (Pseudo-TORCH is another genetic disorder in which children exhibit symptoms similar to those seen in Toxoplasma, Rubella, CMV, HSV infections). Therefore the phenotypic spectrum for the RNase H2 complex mutations is considered to be much broader than classically-defined AGS, and the genetic basis of many cases of "congenital viral infection" may currently be going unrecognised.

The present invention thus provides a polynucleotide comprising the nucleotide sequence of SEQ ID Nos 1 , 3 or 5 or homologs thereof. SEQ ID No 1 gives the nucleotide sequence of AGS2 (RNASEH2B); SEQ ID No

3 gives the nucleotide sequence of AGS3 (RNASEH2C); and SEQ ID No 5 gives the nucleotide sequence of AGS4 (RNASEH2A).

The polynucleotide sequences are reported (with unknown function) in NCBI at NM_024570 (AGS2/RNASEH2B, previously termed FLJ11712); AF312034 (AGS3/RNASEH2C, previously termed AYP1 ); NM_006397 (AGS4/RNASEH2A).

By the term "homolog" or "homologs" with reference to a polynucleotide, we refer to a polynucleotide modified by deletion, substitution or addition of nucleic acids to have at least 65% homology, for example at least 70% homology, for example at least 74% homology to the nucleotide sequence(s) as set out in SEQ ID Nos 1 , 3 or 5. In one embodiment the polynucleotide will have 75% homology or 80% homology or more, preferably 85% homology, to the nucleotide sequence(s) as set out in SEQ ID Nos 1 , 3 or 5 The term "homolog" or "homologs" includes orthologous genes, that is the equivalent gene in a different species.

In one embodiment the homolog will have 90% or more homology, for example 91 %, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% homology, to the nucleotide sequence(s) as set out in SEQ ID Nos 1 , 3 or 5 and when assessed by direct sequence alignment and comparison sequence.

Sequence homology can be determined by direct best-fit sequence alignment and comparison, or by using any suitable homology algorithm, such as BLAST. BLAST is described by Altschul et al., in J MoI Biol 25:403 (1990). The score of an alignment, S, is calculated as the sum of substitution and gap scores. "Substantial homology" when assessed by BLAST refers to low Expectation (E) values. Expectation value is the number of different alignments with scores equivalent or better than S that

are expected to occur in a database search by chance. The lower the E value, the more significant the score.

In particular, modifications to the nucleotide sequence which do not affect the amino acid expressed (due to redundancy in the genetic code) fall within the definition of "homologs". Additionally, the term "homologs" also includes a polynucleotide capable of hybridising to a polynucleotide comprising 15 contiguous bases from any one of SEQ ID Nos 1 , 3 or 5, preferably under stringent conditions. In one embodiment the polynucleotide hybridises to a polynucleotide comprising 20 or more contiguous bases (for example 25 to 50 contiguous bases) from any one of SEQ ID Nos 1 , 3 or 5, preferably under stringent conditions.

Stable hybridisation of polynucleic acids is a function of hydrogen base pairing. Hydrogen base pairing is affected by the degree to which the two polynucleotide strands in the duplex are complementary to each other and also the conditions under which hybridisation occurs. In particular salt concentration and temperature affect hybridisation. One of ordinary skill in the art would be aware that the effective melting temperature (E Tm) of the polynucleotide duplex is controlled by the formula.

E Tm = 81.5 + 16.6(log M [Na ⁺ ]) + 0.41 (% G + C) - 0.72(% formamide)

Where hybridisation is conducted under stringent conditions, only sequences having a high degree of complementary base pairs will remain in duplex form. As used herein the term "stringent conditions" with respect to hybridisation refers to wash conditions of 0.1 X SSC at 60 to 68 ⁰ C. Optionally, the wash conditions can include a suitable concentration of SDS, for example 0.1 % SDS.

The polynucleotide can be DNA or RNA and can be single stranded or double stranded. Double stranded DNA (eg. cDNA) is usually convenient for most applications. The polynucleotide can be in the form of a vector, for example an expression vector.

The polynucleotides of the present invention can be isolated polynucleotides or can be recombinant. The polynucleotides can be incorporated into expression or cloning vectors. Such vectors can be used to transfect or transform host cells and the host cells cultured in conventional culture media according to methods known or described in the art.

Incorporation of cloned DNA into a suitable vector, transfection or transformation of host cells and selection of the transfected or transformed cells are all processes well known to those skilled in the art and numerous suitable methods are described in the literature (see, for example, Sambrook et al., Molecular Cloning: A laboratory Manual, 3 ^rd edition, Cold Spring Harbor Laboratory Press, 2001 ).

Suitable host cells include bacterial, yeast, insect, mammalian and plant cells. Generally the host cell will be selected to be compatible with the vector used.

In one embodiment the host cell can be a human or non-human ES cell.

In a further aspect the present invention provides a lymphoblastoid cell line expressing mutant RNase H2.

By the term "mutant" with reference to a protein (eg. RNase H2A, RNase H2B or RNase H2C) we refer to such a protein in which the amino acid

sequence is different to the wild-type amino acid sequence. For RNase H2B the wild-type sequence is set out in SEQ ID No 2, for RNase H2C the wild-type sequence is set out in the SEQ ID No 4, and for RNase H2A the wild-type sequence is set out in SEQ ID No 6.

The mutant protein can exhibit altered functionality or activity relative to the wild-type protein, but this is not essential. The term "mutant" with reference to a protein complex (eg. RNase H2) refers to the complex in which at least one of the component proteins is a mutant protein as defined above. The mutant protein complex can exhibit altered functionality or activity relative to the wild-type complex, but this is not essential.

In one embodiment, the present invention provides a recombinant polynucleotide comprising the nucleotide sequence as set out in SEQ ID No 1 or homologs thereof, and the protein encoded by that sequence.

In one embodiment, the present invention provides a recombinant polynucleotide comprising the nucleotide sequence as set out in SEQ ID No 3 or homologs thereof, and the protein encoded by that sequence.

In one embodiment, the present invention provides a recombinant polynucleotide comprising the nucleotide sequence as set out in SEQ ID No 5 or homologs thereof, and the protein encoded by that sequence.

In a further aspect, the present invention provides a protein comprising the amino acid sequence of SEQ ID Nos 2, 4 or 6, or homologs thereof.

By the term "homolog" (or "homologs") with reference to a protein, we refer to a protein modified by deletion, substitution or addition of amino

acids to have at least 65% homology, for example at least 66% homology, or for example at least 70% homology to the amino acid sequence as set out in SEQ ID No 2, 4 or 6. In one embodiment the protein will have at least 75% homology or 80% homology, preferably 85% homology, to the amino acid sequence as set out in SEQ ID No 2, 4 or 6. In one embodiment the homolog will have 90% or more homology, for example 91 %, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% homology, to the amino acid sequence as set out in SEQ ID No 2, 4 or 6. The term "homolog" (for "homologs") includes orthologous proteins from different species.

Additionally the present invention provides a protein encoded by the nucleotide sequence of any one of SEQ ID Nos 1 , 3 and 5, or homologs thereof. The nucleotide sequence can be part of a polynucleotide which is optionally a recombinant polynucleotide. The polynucleotide can form part of a larger polynucleotide and, further, can form part of a vector.

In one embodiment the present invention provides a protein comprising the amino acid sequence as set out in SEQ ID No 2 or homologs thereof.

In one embodiment the present invention provides a protein comprising the amino acid sequence as set out in SEQ ID No 4 or homologs thereof.

In one embodiment the present invention provides a protein comprising the amino acid sequence as set out in SEQ ID No 6 or homologs thereof.

In one embodiment the protein of the present invention can form part of a chimeric (fusion) protein.

For clarity the term "protein" is used herein to also refer to a peptide and polypeptide and does not denote any particular size of the polymer.

In a further aspect, the present invention provides a recombinant form of RNase H2 which comprises:

i) a protein encoded by the nucleotide sequence of SEQ ID No 1 , of SEQ ID No 3, or of SEQ ID No 5, or homologs thereof; or ii) a protein comprising the amino acid sequence of SEQ ID No 2, of SEQ ID No 4, or of SEQ ID No 6, or homologs thereof.

In one aspect, the present invention provides a recombinant form of RNase H2 which comprises at least one of:

i) RNASEH2B which is encoded by the nucleotide sequence of

SEQ ID No 1 or a homolog thereof, or which has the amino acid sequence of SEQ ID No 2 or a homolog thereof; ii) RNASEH2C, which is encoded by the nucleotide sequence of

SEQ ID No 3 or a homolog thereof, or which has the amino acid sequence of SEQ ID No 4 or a homolog thereof; or iii) RNASEH2A which is encoded by the nucleotide sequence of SEQ ID No 5 or a homolog thereof, or which has the amino acid sequence of SEQ ID No 6 or a homolog thereof.

In one embodiment at least two of the components i), ii) and iii) are present (eg. i) and iii); i) and ii); or ii) and iii)). In one embodiment all three components i), ii) and iii) are present. Optionally other components of the complex can also be present.

The recombinant RNase H2 complex will have utility in molecular biology, in particular in cDNA production or any other process where DNA-RNA hybrids are to be degraded or suppressed. Also cleavage of duplex DNA containing single or several embedded ribonucleotides. The recombinant RNase H2 complex will also have utility in an assay to identify compounds able to modify the substrate specificity or to modify the activity of the complex. Compounds able to modify the activity of the complex can be activators (agonists) or inactivators (antagonists) to any one of its components or to the complex as a whole. Alternatively compounds able to modify the enzyme activity could act on upstream regulators of RNase H2. The assay can be a cellular assay or a protein assay. Optionally one or more of the components of the complex can be mutated from the wildtype, for example G37S of AGS4. The effect of such a mutation on the activity of the complex or on its substrate specificity can be assessed by assay.

Additionally any one of the components i), ii) or iii) could be used independently in molecular biology or in an assay without one or both of the remaining two components.

RNase H2 is the major source of RNase H activity in the cell and so reduced activity of this enzyme is likely to have consequences for cellular processes dependent on RNA-DNA hybrid metabolism and underlie the pathogenic basis of automimmune diseases including, but not limited to, AGS and SLE, and microbial infection, in particular viral infection. Other microbial infection includes bacterial infection and fungal infection. With reference to viral infection, particular mention may be made of pandemic viral infections, for example pandemic viral influenza where inappropriate inactivation of the innate immune response contributes to high morbidity.

RNase H2 has been proposed to be involved in the removal of Okazaki fragment RNA primers during lagging strand DNA replication. Though deletion of Rnh2 in S. cerevisiae does not effect viability, it does result in increased sensitivity to hydroxyurea.

RNase H2, unlike type 1 RNase H, is also able to recognise single ribonucleotides embedded in DNA, and this enzyme could therefore be important in recognition and processing of inappropriately incorporated ribonucleotides in genomic DNA. Such misincorporation might occur more frequently in circumstances of depleted dNTPs, and provide an alternative explanation for hydroxyurea sensitivity of Rnh2 mutants.

When single stranded RNA associates with one strand of duplex DNA, the opposite DNA strand is displaced forming an "R-loop" of single stranded DNA. Such loops can occur following transcription in eukaryotic cells when RNA binding proteins are disrupted. RNase H appears to have a role in suppressing such structures and the resulting genomic DNA instability.

Therefore, abrogation of RNase H activity may have consequences for

DNA synthesis, removal of misincorporated ribonucleotides in DNA, and/or suppression of R-loop formation.

Reverse transcription is an essential process for the replication of HIV and other retroviruses and is an important drug target for antiviral therapy. DNA-RNA hybrids formed during this process will be susceptible to degradation by endogenous RNase H, and these could therefore have an important antiviral role. Thus, mutation of RNase H2 may impair host antiviral defences, and AGS could be the consequence of common viral pathogens.

Alternatively immune misregulation may be the basis for AGS, SLE and may be significant in autoimmune disease and microbial infection, if reduced cellular RNase H activity increases endogenous levels of RNA- DNA hybrids. dsRNA and dsDNA are known activators of innate immunity and activate type I interferon production (see Kawai et al., Nat Immunol 7:131-137, 2006 and Krieg et al., Annu Rev Immunol 20:709-760, 2002). Defects in other nucleases, resulting in reduced clearance of extracellular nucleic acids released from apoptotic cells, are proposed to trigger multisystem autoimmune diseases as a consequence of circulating nucleic acids. Supporting evidence includes the finding of heterozygous mutations in DNase 1 in several human SLE patients and DNase1 ^"λ mice exhibiting a lupus-like phenotype (see Napirei et al., Nat Genet 25:177- 181 , 2000). Furthermore DNasell ^" ' ^" IfnR ^" ' ^" mice have recently been reported to exhibit chronic polyarthritis resembling rheumatoid arthritis (see Kawane et al., Nature 443:998-1002, October 2006), and it is believed that RNA-DNA hybrids may also be a stimulus for innate immunity, and explain the high levels of interferon alpha that are a diagnostic feature of AGS. RNase H2 dysfunction could similarly raise levels of endogenous RNA-DNA hybrids which then stimulate interferon alpha production by mechanisms comparable to those for dsRNA and dsDNA. Consequent overproduction of interferon alpha in the central nervous system could explain many of the neuropathological features of AGS, as shown by a transgenic mouse model, in which interferon alpha was chronically expressed in glial cells (see Campbell et al., Brain Res 835:46-61 , 1999).

The link between SLE and autoimmune disease is demonstrated by recent studies, for example Kelly et al., Arthritis Rheum 54:1557-1567, 2006 showed that binding of dsRNA to TLR causes IFN production; Lovgren et al., Arthritis Rheum 50:1861-1872, 2004 showed that anti-RNA Abs are

needed for IFN induction in an in vitro SLE assay; Sigurdsson et al. Am J. Hum. Genet. 76:528-537, 2005 showed that SLE is based on a dysfunction of the IFN pathway and Graham et al., Proc. Natl. Acad. Sci USA 104:6758-6763, 2007 showed mutations in IFN regulatory factor is associated with increased risk of SLE.

The innate immune response in influenza is shown to have a common aetiology with SLE by recent teachings. For example Cheung et al., Lancet 360:1831-1837, 2002 showed increased TNF alpha is associated with the elevated pathology of H5N1 (97); Kobasa et al., Nature 445:319- 323, 2007 showed an atypical innate immune response to 1918 flu in non- human primates and Lipatov et al., Journal of General Virology 86:1121- 1130, 2005 showed nine models infected with H5N1 flu show cytokine imbalance.

It would therefore be useful to provide compounds able to modify the activity or substrate specificity of RNase H2 in order to:

a) increase RNase H2 activity to combat the symptoms of AGS. b) increase RNase H2 activity to combat microbial infection, in particular, but not limited to, bacterial and viral infection. Particular mention may be made of retroviral infection (eg. HIV infection) where DNA-RNA hybrids are formed in the host cell during viral replication. Also of pandemic strains of influenza where inappropriate activation of the innate immune response has been implicated in the high morbidity of such strains e.g. 1918 and H5N1 strains (see Cheung et al., The Lancet 360:1831-1837, 2002 and Lipatov et al., Journal of General Virology 86:1121-1130, 2005).

c) increase RNase H2 activity to decrease genomic DNA instability, in particular by suppressing R-loop formation. d) increase RNase H2 activity to combat autoimmune disease. Exemplary autoimmune diseases include (but are not limited to) Systemic Lupus Erythmatosis (SLE), coeliac disease, Crohn's disease, Diabetes mellitus (type 1 ), Goodpasture's syndrome, Graves' disease, Addison's disease, rheumatoid arthritis, psoriasis, and multiple sclerosis. e) increase RNase H2 activity to enhance the efficiency of the complex in molecular biology reactions, for example in the production of cDNA. f) decrease RNase H2 activity to promote the efficacy of viral- based gene therapy. (RNase H2 activity would form part of the host response which impairs the utility of this method. A reduction in RNase H2 activity would at least partially suppress this host response). g) decrease RNase H2 activity to stimulate the host immune response to viral infection (see Akwa et al., J. Immunol 161 :5016-26, 1998).

The present invention further provides an assay to identify an activator or inactivator of RNase H2 or a component thereof, or a modulator of RNase H2 substrate specificity, said assay comprising: i) contacting a test substance with RNase H2 or a component thereof wherein said component is RNASEH2B, RNASEH2C,

RNASEH2A or any combination thereof; ii) assessing the activity of RNase H2 or the component thereof in the presence of the test substance, optionally over a period of time; and

iii) determining the effect of the test substance on activity of RNase H2 or the component thereof.

The assay can be directed to the whole RNase H2 complex, which may be a recombinant complex as defined above. Alternatively, the assay could be modified to be directed to component(s) of the RNase H2 complex, for example RNASEH2B, RNASEH2C or RNASEH2A each of which can be recombinant.

The assay described above can also be performed by contacting the test substance with a mutant form of RNase H2 or a mutant component thereof (eg mutant RNASEH2A, RNASEH2B or RNASEH2C, or combinations thereof). The mutant form of RNaseH2 or its component(s) can be expressed by a cell line such as a lymphoblastoid cell line or by a non- human transgenic animal, such as a rodent.

In one embodiment the RNaseH2 complex is present within a cell.

In one embodiment the assay can be directed to identify an inflammatory modulator, that is an anti-inflammatory or pro-inflammatory agent. In one embodiment the assay can be directed to identify an anti-microbial (eg. anti-viral) agent.

The assay can be a cellular based assay where endogenous RNase H2 activity is assessed. The assay could identify modulators acting in an indirect way on RNase H2 or on a pathway upstream of RNase H2. The assay will be useful in assessing the effect of inactivators or inhibitors of viral RNase H on mammalian (eg. human) RNase H2 as such inactivators or inhibitors, if specific to viral RNase H alone, could form an effective anti-viral treatment. The impact of such putative anti-viral agents

on the RNase H2 complex described here would be critical in determining their efficacy in vivo, and in particular the likely extent of side-effects on the mammalian host.

Step ii) of the assay described above can be achieved, for example, by introducing a labelled oligonucleotide substrate. In one embodiment the labelled oligonucleotide when cleaved by RNase H2, releases a fluorescent tag. The intact oligonucleotide does not exhibit fluorescence due to the proximity of a quencher molecule attached to the opposite nucleotide strand. Thus the degree of fluorescence is a measure of RNase H2 activity. In this embodiment the oligonucleotide acts as a substrate for the RNase H2 or component thereof. The oligonucleotide can be double-stranded DNA, doublestranded RNA or a double -stranded DNA-RNA hybrid molecule.

Step iii) may be achieved by comparing RNase H2 activity under identical conditions, but in the absence of any test substance.

The assay can be conducted over a period of time (ie. can be a kinetic assay), for example over a time period of 10 to 60 minutes. The activity of the RNase H2 could be assessed at suitable (preferably regular) time points throughout that time period, for example from every 2 minutes to every 20 minutes. In one embodiment the assay is conducted over a period of 30 minutes with activity measured every 5 minutes.

The present invention further provides an assay to determine the effect of a modification (for example an amino acid mutation) in one or more of the components of RNase H2, said assay comprising:

i) providing RNase H2, wherein at least one component thereof is a mutant relative to the wild type form thereof; ii) contacting said RNase H2 with a substrate therefor; iii) assessing the activity of the RNase H2, optionally over a period of time; and iv) determining the effect of the modified component on RNase H2 activity.

In one embodiment the wildtype form of RNASEH2B has the amino acid sequence of SEQ ID No 2. In one embodiment the wild-type form of RNASEH2C has the amino acid sequence of SEQ ID No 4. In one embodiment the wild-type form of RNASEH2A has the amino acid sequence of SEQ ID No 6.

In one embodiment the RNase H2 is recombinant.

In one embodiment the RNase H2 is expressed from a cell line, such as a lymphoblastoid cell line. Alternatively the RNase H2 can be expressed by a non-human transgenic animal.

In one embodiment the RNaseH2 complex is present within a cell.

Step ii) of the assay described above can be achieved, for example, by introducing a labelled oligonucleotide substrate. In one embodiment the labelled oligonucleotide when cleaved by RNase H2, releases a fluorescent tag. The intact oligonucleotide does not exhibit fluorescence due to the proximity of a quencher molecule attached to the opposite nucleotide strand. Thus the degree of fluorescence is a measure of RNase H2 activity. In this embodiment the oligonucleotide acts as a substrate for the RNase H2 or component thereof. The oligonucleotide

can be double-stranded DNA with embedded ribonucleotide(s), RNA or a double-stranded DNA-RNA hybrid molecule.

Step iii) may be achieved by comparing RNase H2 activity under identical conditions, but in the absence of any test substance.

In one embodiment, a modified component of the RNase H2 complex is assayed in the presence or absence of the other components. In such an embodiment, it may be useful for a mutant form of the RNase H2 component which affects functionality to be used, for example RNASEH2A (AGS4) G37S or RNASEH2B (AGS2) A177T, and the assay conducted will look for re-establishment of RNase H2 activity. Alternatively the assay conducted will look for reduction of RNase H2 activity. Alternatively the assay will look for a change in substrate specificity of the RNase H2.

The assay can be conducted over a period of time (ie. can be a kinetic assay), for example over a time period of 10 to 60 minutes. The activity of RNase H2 can then be assessed at suitable (preferably regular) time points throughout that time period, for example from every 2 minutes to every 20 minutes. In one embodiment the assay is conducted over a period of 30 minutes with activity measured every 5 minutes.

An RNASEH2 knock-out mouse can be made using standard techniques. Such mice could be challenged with virus as described for a transgenic mouse expressing IFN-α as in Akwa et al., J Immunology 161 :5016-5026 (1998) and then used to test agonists of mammalian RNase H2, optionally identified by the assay above, in vivo to see if the effects of the virus are ameliorated. Such a knock-out mouse forms a further part of the present invention.

In a further aspect, the present invention provides an assay to detect mutations in the RNASEH2B, RNASEH2C or RNASEH2A genes, in the genome of a patient, said assay comprising: i) causing lysis of white blood cells obtained from the patient; and ii) assessing RNase H2 activity of the lysed white blood cells.

The mutations detected will be mutations that affect RNase H2 activity, either by suppressing or enhancing its activity relative to the normal range of the wild-type complex.

Optionally, where decreased or increased RNase H2 activity relative to the normal range is detected, sequencing of the RNASEH2B, RNASEH2C or RNASEH2A gene(s) can be conducted.

The assay would identify patients exhibiting decreased RNase H2 activity due to homozygous or heterozygous mutations in any of RNASEH2B, RNASEH2C or RNASEH2A as an aid for diagnosis of AGS or so that appropriate counselling could be provided to a patient at risk of having children with AGS . The assay could also be useful to identify individuals with autoimmune disease, congential viral infection or increased viral infection susceptibility (due to reduced levels of RNase H2 activity) to allow diagnosis and therapy.

The assay could likewise also identify patients with SLE or other autoimmune disease or at risk of having children with SLE or other autoimmune disease. The assay could further provide information on the extent of a viral infection, or be used as an aid to diagnosis for viral infection, or to identify patients with an increased susceptibility to microbial infection, in particular, but not limited to, viral infections.

In an alternative embodiment, the assay to determine RNase H2 activity could simply be conducted by genotyping a genetic sample provided by the patient, either by fingerprint or satellite techniques or by sequencing of the relevant portion of the genome. The genetic information provided could be used to identify patients having altered RNase H2 activity as an aid to diagnosis of AGS, SLE, autoimmune disease, increased susceptibility to microbial infection, or to identify hereditary risk or predisposition thereto.

In a further aspect, the present invention includes polyclonal or monoclonal antibodies able to bind specifically to a protein encoded by the nucleotide sequence of any one of SEQ ID Nos 1 , 3 or 5 or homologs thereof; a protein comprising the amino acid sequence of any one of SEQ ID Nos 2, 4 or 6 or homologs thereof; or a recombinant RNase H2 complex as described above. Antibodies to mutated forms of any one of the RNase H2 complex components (eg. RNASEH2A, G37S mutation) are also included. Monoclonal antibodies can be generated, for example by immunising mice with enzymatically active recombinant RNase H2 complex and subsequently performing hybridoma fusion using known methodologies. Clones can be screened by ELISA, and further validated using immunofluorescence and western blot assays using cells expressing epitope-tagged constructs. The antibodies can be used to aid molecular biology procedures, to purify RNase H2 (for example in diagnostic tests), to aid study of the RNase H2 complex and its function.

In one embodiment, the antibodies are therapeutic antibodies, and are produced from a cell line, such as a hybridoma (see Kohler et al., Nature 256:495-497, 1975 and Galfre Meth Enzymol 73:3, 1981 ). Suitable therapeutic antibodies include murine antibodies, chimeric antibodies, scFvs, humanized antibodies and human antibodies. Chimeric antibodies

are genetically engineered antibodies containing approximately one third non-human protein and approximately two thirds human protein. A humanized antibody is genetically engineered to contain the minimum amount non-human protein (typically 5 to 10%) on a human antibody to minimise adverse host immune reaction thereto (see Riechmann et al., Nature 332:323-327, 1988).

Fragments of antibodies, such as F(ab')2, Fab and Fv fragments, are also included by the term "antibody".

The present invention also provides a diagnostic kit containing at least one primer for RNASEH2A, RNASEH2B or RNASEH2C. Suitable primers include those set out in Table 1 , in combination with suitable auxiliaries. Suitable auxiliaries, as used herein include buffers, dNTPs, enzymes (eg. TAQ polymerase), labelling compounds and the like. The kit will be used to look for genetic aberrations in the sample nucleic acid, such aberrations (if found) can indicate a genetic disposition to AGS, SLE, autoimmune disease or an elevated susceptibility to microbial infection, in particular, but not limited to viral infections.

In one embodiment an antibody specific to the RNase H2 complex as described above can be used to identify RNase H2 containing a mutant protein.

In an alternative embodiment, an antibody specific to the RNase H2 complex as described above can be used in therapy. For example the antibody can be used to promote the efficacy of viral-based gene therapy or to stimulate the host immune response.

The present invention provides a method of producing a transgenic non- human animal having a genome encoding mutant RNASEH2A, RNASEH2B or RNASEH2C, said method comprising: a) introducing a recombinant genetic construct comprising a polynucleotide sequence encoding mutant RNASEH2A,

RNASEH2B or RNASEH2C into a non-human zygote or a non- human embryonic stem cell; b) generating a transgenic non-human animal from said zygote or embryonic stem cell; and c) producing a transgenic non-human animal having a genome encoding mutant RNASEH2A, RHASEH2B or RNASEH2C.

The polynucleotide sequence encoding mutant RNASEH2A, RHASEH2B or RNASEH2C will preferably be linked to a promoter so that the protein is expressed by the transgenic animal.

The non-human transgenic animal having a genome encoding mutant RNASEH2A, RHASEH2B or RNASEH2C forms a further aspect of the present invention.

The non-human transgenic animal can be any suitable animal and mention is made of mice, rats, primates, hamsters, rabbits, dogs, cats, fish, cattle, swine and sheep as suitable examples. However, this list is not exhaustive and other animals could also be used.

In one embodiment the non-human transgenic animal is a rodent. Suitable examples include rats and mice.

Embryonic stem cells used in the art which may be used in the methods of this invention comprise but are not limited to embryonic stem cells derived

from mouse strains such as C57BL/6, CBA/, BALB/c, DBA/2 and SV129. Preferably, embryonic stem cells derived from C57BL/6 mice are used (Seong, E et al., Trends Genet. 20, 59-62, 2004; Wolfer et al., Trends Neurosci. 25:336-340, 2002).

A transgene (including the coding sequence of RNASEH2A, RNASEH2B or RNASEH2C can be introduced into the germline of an animal using a variety of methods. For example, the transgene can be directly injected into the male pronucleus of a fertilized egg (see, e.g., Hogan et al., Manipulating the Mouse Embryo, Cold Spring Harbor Laboratory, Cold Spring Harbor Press (1994)), resulting in the random integration into one locus of a varying number of copies of the transgene, usually in a head to tail array (see for example Costantini and Lacy, Nature, 294: 92, 1981 ). The injected eggs are then re-transferred into the uteri of pseudopregnant recipient mothers. Some of the resulting offspring may have one or several copies of the transgene integrated into their genomes, usually in one integration site. These "founder" animals are then bred to establish transgenic lines of animals and to back-cross into the genetic background of choice. One of ordinary skill in the art will appreciate the advantage of introducing the transgene on both chromosomes. Alternatively, transgenes can be introduced into an animal by gene targeting, for example by homologous recombination, in embryonic stem (ES) cells. Suitable known methods for introducing the transgene include blastocyst injection. In this regard, a targeting construct comprising the transgene is prepared using methods known in the art.

A target construct comprising the transgene can be introduced in a suitable host cell using methods known in the art. Production of transgenic embryos and their screening can be performed using known techniques, for example as described by Joyner ed., Gene Targeting, A

Practical Approach, Oxford University press, 1993. The DNA of the transgenic animal or embryo can be screened by southern blot or by PCR techniques.

The transgenic non-human animal can be a healthy animal or may exhibit symptoms of a disease or disorder caused by a mutation in the transgene. Such transgenic animals are well suited for pharmacological studies of drugs and can be used as disease models, for example of AGS, SLE, autoimmune disease or viral infection. Such models may have an elevated susceptibility to AGS, SLE, autoimmune disease or viral infections.

The present invention will now be further described with reference to the following, non-limiting, examples, and figures in which:

Figure Legends

Figure 1 Neuroimaging and clinical findings in Aicardi-Goutieres syndrome, (a) Axial CT scan showing calcification of the basal ganglia, (b) Axial T2 MRI demonstrating high signal intensity in the white matter, particularly affecting the frontal lobes. For comparison, (c) Axial CT scan of patient with congenital HIV infection showing calcification of the basal ganglia, from Belman et al., Neurology 36: 1192-1199 (1986).

Figure 2 Schematic of AGS2 critical region and RNASEH2B (previously termed FLJ11712) gene depicting location of identified mutations, (a)

Genetic map of chromosome 13q14.1 depicting the refined AGS2 locus, defined by overlapping homozygous chromosomal segments in non- consanguineous families, and extending between microsatellite markers AC137880TG19 and D13S788. (b) The physical map of the 571 kbp critical interval contains 4 annotated genes (UCSC Genome Browser May 2004

assembly), (c) RNASEH2B (FLJ11712) spans 47kbp of genomic sequence in 11 exons, and encodes a 308 amino acid protein. Coding sequence shaded grey. Locations of mutations are indicated by arrows, with mutation position enumerated relative to the translational start site, and the corresponding amino acid change in bold.

Figure 3 Schematic of AGS3 region, the RNASEH2C (AYP1) gene, its mutations and sequence conservation in other species, (a) Genetic map of chromosome 11 q13.1. The critical region is defined by linkage disequilibrium data in the Pakistani families and lies between D11S4205 and D11S987. The RNASEH2C (AYP1) gene lies at 65.2 Mbp spanning 1.4kbp of genomic sequence in a telomeric to centromeric orientation on the minus strand, (b) The RNASEH2C (AYPI) gene structure comprises 4 exons (coding sequence in grey), and encodes a 164 amino acid protein. The location of mutations are indicated by arrows, with mutation position enumerated relative to the translational start site, and the corresponding amino acid change in bold, (c) Both mutations lie in residues conserved in mammals. Sequence alignments of regions of mammalian RNASEH2C proteins that immediately flank the mutations. Hs, Homo sapiens; Bt, Bos Taurus; Cf Canis familiaήs; Mm, M us musculus; Rn, Rattus norvegicus. Substituted amino acids are shown above sequence alignment, (d, e) Sequence electropherograms of RNASEH2C (AYP1) mutations, (d) c.428A>T (e) C.205OT.

Figure 4 The RNASEH2A gene, genomic location, gene structure and mutation location, (a) Genetic map of chromosome 19p13.13.

RNASEH2A lies in a region, where 2 affected children of consanguineous 2 ^nd cousin parents (family F39) share homozygous SNPs (SNP A- 1509361 , 1606327, 1606325), defining a region of potential genetic linkage between SNP A-1515950 and A-1508018. (b) RNASEH2A gene structure. RNASEH2A lies at 12.8Mbp (UCSC Genome Browser May 2004

assembly), spanning 7kbp of genomic sequence. It comprises 8 exons and encodes a 299 amino acid protein, (c) The G37S mutation occurs at a residue which is absolutely conserved from bacteria to humans. Hs, Homo sapiens; Cf, Canis familiaris; Mm, M us musculus; Rn, Rattus norvegicus; Ce, Caenorhabditis elegans; Sp, Schizosaccharomyces pom be; Sc, Saccharomyces cerevisiae] Ec, Escherichia coli; Ph, Pyrococcus horikoshii. (d) Electropherogram of RNASEH2A mutation (e) Predicted tertiary structure of RNASEH2A catalytic site modelled on solved crystal structures of type 2 RNase H proteins. The mutated G37 residue (centre) lies in close proximity to the active site and putative substrate binding residues (Chapados et al., J MoI Biol 307:541 -556 (2001 )).

Figure 5 Human RNASEH2B (FYJ11712), RNASEH2C (AYP1 ) and RNASEH2A form an enzymatically active Type Il Ribonuclease H complex when expressed in mammalian cells, (a) Schematic representation of the proposed human RNASEH2 complex and its S. cerevisiae counterpart, (b) T7 Epitope tagged FYJ 11712 and AYP1 can be co-immunoprecipitated (IP) with myc-tagged RNASEH2A. (c, d) The RNASEH2A/B/C complex exhibits ribonuclease H activity, (c) Three oligonucleotide heteroduplexes were used to assay for enzymatic activity. Oligo C, a substrate degradable by any ribonuclease H, is a 3'-fluorescein tagged RNA oligonucleotide hybridised with a complementary 5' labelled DABCYL DNA oligonucleotide (DABCYL quenches the fluorescein until enzymatic cleavage with consequent dissociation of the fluorescently labelled 3'fragment). Oligo B, is a substrate degradable only by Type Il ribonuclease H. It is a DNA duplex with a single ribonucleotide at position 15 in the oligonucleotide strand that is 3'-labelled with fluorescein. Oligo A, is a RNA:DNA hybrid, like oligo C, but enzymatically resistant, as the 3'fluorescein labelled oligonucleotide has been synthesised with 2'O methyl RNA nucleotides.

(d) Type Il Ribonuclease H activity can be immunoprecipitated from HEK293T extracts containing epitope-tagged RNASEH2A/B/C. Myc IP, performed with mouse anti-myc antibody; IgG IP, control immunoprecipitation performed with normal mouse IgG immunoglobulin. Vector, cells transfected with empty pCGT-Dest and pcDNA3.1 mychis vectors. Error bars, s.e.m.

Figure 6 Mutation in RNASEH2A reduces RNase H activity, (a) Immunoprecipitation of the RNase H2 complex from HEK293T cells co- transfected with a mutant form of the RNASEH2A protein, in combination with wild-type tagged RNASEH2B and RNASEH2C, showing that the RNASEH2A G37S mutation does not appear to disrupt complex formation. In, input, IP, immunoprecipitate, WT, wild-type, (b) Mutation in the RNASEH2A reduces enzyme activity. Fluorometric RNase H assay of the same immunoprecipitated complexes shown in (a). Error bars, s.e.m.

Figure 7 Microsatellite genotyping in two non-consanguineous families refines the AGS2 critical interval. Regions of homozygous markers boxed.

Figure 8 Multiple sequence alignment of (a) RNASEH2B/Rnh2Bp (b)

RNASEH2C/Rnh2Cp homologues from representative eukaryotic species. Amino acids substituted in AGS patients are shown in white-on-black and their substituting amino acids are given above. The Kluyveromyces waltii amino acid sequence used to establish homology between human AYP1 and S. cerevisiae Rnh2Cp is included in (b). Alignments are presented using CHROMA (Goodstadt et al., Bioinformatics 17:845-846 (2001 )) and an 80% consensus of the eukaryotic sequences. Residues which have been excised from the alignment are indicated in parentheses and spacers (-) are used for alignment purposes only and do not represent any deletion. Genlnfo identifiers are presented to the right of the alignments.

Consensus abbreviations (amino acids): a, aromatic (FHWY); b, big (EFHIKLMQRWY); C, charged (DEHKR); h, hydrophobic (ACFGHILMTVWY); I, aliphatic (ILV); p, polar (CDEHKNQRST); s, small (ACDGNPSTV); *, Ser/Thr (ST); +, positively-charged (HKR); and -, negatively-charged (DE). Species abbreviations: Ag, Anopheles gambiae; Am, Apis mellifera; An, Aspergillus nidulans; Bt, Bos taurus; Ca, Candida albicans; Ce, Caenorhabditis elegans; Cf, Canis familiaris; Cg, Candida glabrata; Cp, Cryptosporidium parvum; Dd, Dictyostelium discoideum; Dh, Debaryomyces hansenii; Dm, Drosophila melanogaster, Dr, Danio rerio; Eg, Eremothecium gossypii; Gz, Gibberella zeae; Hs, Homo sapiens; Kl, Kluyveromyces lactis; Kw, Kluyveromyces waltii; Mm, Mus musculus; Nc, Neurospora crassa; Os, Oryza sativa; Rn, Rattus norvegicus; Sb, Saccharomyces bayanus; Sea, Saccharomyces castellii; See, Saccharomyces cerevisiae; SkI, Saccharomyces kluyveri; Sku, Saccharomyces kudriavzevii; Sm, Saccharomyces mikatae; Spa,

Saccharomyces paradoxus; Spo, Schizosaccharomyces pombe; Tb, Trypanosoma brucei; Tc, Trypanosoma cruzi; Tn, Tetraodon nigroviridis; Xt, Xenopus tropicalis; and, Yl, Yarrowia lipolytica.

Figure 9 The >4GS3 locus maps to chromosome 11q13.2. Microsatellite genotyping in six Asian families. Postulated ancestral haplotype shown in bold, and regions of homozygous markers boxed.

Figure 10 RNase H activity assay in polyinosine-polycytidic acid (poly(l:C)) treated HCT116 cells. A: enzyme resistant oligonucleotide substrate, B: RNaseH2 specific substrate, C: Ribonuclease H substrate.

Figure 11 Western blot using the affinity purified polyclonal antibody raised against the anti-RNase H2 A/B/C complex from sheep against HeK293T cell lysate in which the RNase H2 A/B.C complex proteins have

been overexpressed using epitope tagged vectors. The antibody is able to detect over-expressed tagged mammalian RNase H2A, H2B and H2C (both lower bands). ^* is a putative endogenous band.

Figure 12 Schematic diagram of RNASEH2B ^A177T targeting construct introduced to ES cells. Targeting construct contains 2.5kb arms of homologous sequence from the RNASEH2B genomic locus, with PGK- neomycin cassette inserted between exons 6 and 7, flanked by loxP sites (triangles). A nucleotide change (indicated by arrow) has been introduced to create the common AGS patient Alanine to Threonine mutation at codon 177. This construct has been used for successful homologous recombination to create an ES cell line that is currently being used for blastocyst injections to create a "knock-in" mouse model of AGS.

Figure 13 Kinetic RNase H2 activity assay on immortalised AGS patient lymphoblastoid cell lines for wild type LCL (WT LCL), RNase H2A G37S, and RNase H2B A177T. Cell lines containing the indicated homozygous mutations in the RNase H2 subunits, have reduced enzyme activity, relative to a wild-type lymphoblastoid cell line (WT).

Figure 14 Endpoint fluorescent RNase H assay showing AGS patient lymphoblastoid cell lines have reduced RNase H2 enzyme activity.

Figure 15 a. SDS PAGE gel of soluble recombinant RNase H2 protein complex; b. Enzymatic activity of recombinant GST, wild-type GST-RNase H2 complex and GST-RNase H2 complex with the G37S mutation in the A subunit.

Examples

Example 1 : Patients and subjects

All affected individuals included in the study fulfilled diagnostic criteria for AGS, with neurological features of an early onset encephalopathy, negative investigations for common prenatal infections, intracranial calcification in a typical distribution, a CSF lymphocytosis >5 cells/mm ³ and / or >2 IU/ml of IFN-α in the CSF. With consent, blood samples were obtained from affected children, their parents, and unaffected siblings. Genomic DNA was extracted from peripheral blood leukocytes by standard methods. The study was approved by the Leeds Health Authority/United Teaching Hospitals NHS Trust Research Ethics Committee, and by the Scottish Multicentre Research Ethics Committee (04:MRE00/19).

Genotyping and Linkage Analysis

Genome-wide scans by SNP array were performed using Affymetrix Human Mapping10K Xba142 2.0 GeneChips® by MRC Geneservice (Cambridge, UK). High-density genotyping of AGS2 and AGS3 loci was performed as previously described (Jackson et al., Am J Hum Genet 63: 541-546 (1998)) using established microsatellite markers from the

Marshfield linkage maps (http://research.marshfieldclinic.org/genetics), and novel microsatellites identified from the Human Genome Browser sequence (May 2004 freeze, http://genome.ucsc.edu/). Linkage analysis was performed using GENEHUNTER (version 2.0β) (Kruglyak et al., Am J Hum Genet 56: 1347-1363 (1996)) under a model of autosomal recessive inheritance with full penetrance with a disease allele frequency estimated at 1 in 1000 and assuming equal marker allele frequencies.

Bioinformatics

Database searches employed PSI-BLAST

(http://www.ncbi.nlm.nih.gov/blast/; Altschul et al Nucleic Acids Res. 25:3389-3402 (1997), the non-redundant protein sequence database and an E-value inclusion threshold of 2x10 ^"3 .

Comparative protein structure modelling was performed utilising SWISS- MODEL (Schwede et al Nucleic Acids Res 31 : 3381-3385 (2003) using default settings and accuracy of the RNASEH2A protein (NP_006388) predicted structure validated by WHAT_CHECK (Hooft et al Nature 381 : 272 (1996). Four homologous archaeal RNase H2 proteins of known structure (demonstrating 35-40% sequence identity with NP_006388) were employed as best available templates for the prediction, 1 uaxA 1 io2A 1x1 pA and 1ekeB. The active site and predicted substrate contact residues (Chapados et al., J MoI Biol. 307: 541-556 (2001 )) were annotated using VMD (v1.8.3) (Humphrey et al., J MoI Graph 14: 33-38, 27-28 (1996)).

Mutation detection Primers were designed to amplify the coding exons of RNASEH2B,

RNASEH2C and RNASEH2A (primer sequences are set out in Table 1 ). Purified PCR amplification products were sequenced using dye terminator chemistry (Applied Biosystems) and electrophoresed on an ABI 3700 capillary sequencer (Applied BioSystems), or Megabace 500 (Amersham Pharmacia) capillary sequencers. Mutation analysis was performed using Mutation Surveyor (Softgenetics).

Table 1

Vector construction

The Gateway Vector system (Invitrogen) was used for constructing mammalian expression vectors. The coding sequence of RNASEH2B, RNASEH2C and RNASEH2A was amplified from plasmid clones (CSODF031YM15, CR602872, Invitrogen and clone IRAUp969G0361 D, BC011748, RZPD, respectively) and cloned into pDONR221™. AYP1 was purchased as a ready made pENTRY vector (clone IOH27907, Human Ultimate™ Full ORF Gateway Shuttle Clone, NM_032193, Invitrogen). pcDNA3.1 mychis-Dest and pCGT-Dest were constructed

using the Gateway Vector conversion system by insertion of Gateway reading frame cassettes into the multicloning sites of pcDNA3.1 mychis (Invitrogen) and pCGT (Van Aelst et al., Embo J 15:3778-3786 (1996)). Site-directed mutagenesis was performed on the RNASEH2A pENTRY clone using the Stratagene Quikchange kit according to the manufacturer's instructions.

lmmunoprecipitation and Western blotting

HEK293T cells were transiently co-transfected with 1μg of each construct using Lipofectamine (Invitrogen) according to manufacturer's instructions. After 24 hours, cells were lysed in 5OmM Tris (pH 7.8), 28OmM NaCI, 0.5% NP40, 0.2mM EDTA, 0.2mM EGTA, 10% glycerol, 0.1 mM sodium orthovanadate, 1μM DTT and 1μM PMSF for 10min at 4°C. Lysed cells were then diluted 1 :1 with 2OmM Hepes (pH 7.9), 1OmM KCI, 1 mM EGTA, 10% glycerol and 0.1 mM sodium orthovanadate buffer for a further 10min and extracts were cleared by centrifugation (15800 x g, 10min, 4°C). 500μg of protein lysate were immunoprcipitated using Protein A/G PLUS agarose (Santa Cruz), following the manufacturer's protocol, using 1μg of mouse anti-myc antibody (clone 9B11 , Cell Signalling) or 1 μg of mouse IgG (Santa Cruz). Western blots were performed using mouse anti-myc monoclonal antibody at 1/1000 and mouse monoclonal anti-T7 antibody (Novagen) at 1/5000.

RNaseH assays 10μM oligonucleotides (Eurogentec) were annealed in 6OmM KCI, 5OmM TrisHCI pH8 by denaturation at 95 ⁰ C for 5 minutes and then gradual cooling to room temperature. Fluorometric RNase H assays were performed in 100μl volume of 6OmM KCI, 5OmM TrisHCI pH8, 1OmM MgCb, 0.25μM oligonucleotide duplex, in 96 well flat-bottomed plates, at 37 ⁰ C for 3hrs in an orbital shaker at 60 rpm. 1/1 Oth volume of each

immunoprecipitate was used in each reaction, and as a positive control 2.5 units of E.coli RNase H (Invitrogen) was used. Fluorescence was read using a VICTOR ² 1420 multilabel counter (Perkin Elmer), with an 480nm excitation filter, and 535nm emission filter for 100msec.

Results

AGS2 locus refinement and identification of RNASEH2B (FLJ11712) High density genotyping of microsatellite markers was performed in a panel of 10 families to refine the AGS2 locus. Two of the non- consanguineous families (F8 and F10, Figure 7) were found to exhibit small overlapping regions of homozygous markers. Such regions are likely to represent autozygous chromosomal segments (Broman et al., Am J Hum Genet 65:1493-1550 (1999) and Gibson et al., Hum MoI Genet 15: 789-795 (2006)) which in rare autosomal recessive disorders have a high probability of containing the disease gene (Lander et al., Science 265:2049-2054 (1987)). Using this overlapping region of homozygous markers the AGS2 critical region was refined to a 571 kbp region on chromosome 13q14.3 between genetic markers AC1378890TG19 and D13S788 (Fig 2). The coding exons of all four annotated genes in the critical region were sequenced.

RNASEH2B (FLJ11712) is the AGS2 gene

Missense sequence changes in FLJ 11712 were identified in 7 of the families screened, while no pathogenic mutations were evident in DLEU7, GUCY1B2 or FLJ30707. Subsequent mutation screening of FLJ11712 in a larger cohort identified a total of 18 families with mutations in this gene (Table 2, Fig 2). Most mutations were missense, with two being found recurrently in different ethnic groups (A177T, V185G). All missense mutations resulted in non-conservative replacement of residues that are

conserved among mammals (Figure 8), with the exception of a single conservative residue change (Y219H) which is a residue conserved back to Dictostelium. Nonsense mutations were identified in two families, a stop codon in exon 2 (F17), and a splice donor site mutation in intron 6 (F15). In both these cases affected individuals were compound heterozygotes, with the second mutation being a missense change. The observed mutation spectrum therefore suggests that the mutational consequences will be hypomorphic rather than complete loss of FLJ 11712 protein function. Mutations segregated with the disease in all families, and all available parents were heterozygous for the mutations. At least 160 control alleles were genotyped for each mutation. Only for the most common mutation, A177T, was one heterozygous individual found in 241 samples tested.

Table 2 - FLJ11712 mutations identified in 18 Aicardi-Goutieres syndrome families

Family Ethnicity Nucleotide alterations Amino acid Exon Segregati Parental alterations on consanguinity

F1 Moroccan c.485A→C K162T 6 Horn, M, P Yes

F3 Italian c.554T→G V185G 7 Horn, M, P Yes

F4 Algerian c.529G→A A177T 7 Horn, M, P Yes

F5 Moroccan c.529G→A A177T 7 Horn, M, P Yes

F8 Italian c.554T→G V185G 7 Horn, M, P Yes

F9 Irish c.529G→A A177T 7 Horn, M, P No

F10 Moroccan c.529G→A A177T 7 Horn, nps No

F11 Mixed c.[257A→G]+[529G→A] [H86R]+[A177T] 4; 7 Het, M, P No

Tunisian/Algerian

F12 Italian c.[488C→T]+[529G→A] [T163I]+[A177T] 6; 7 Het, P No

F13 Italian c.529G→A A177T 7 Horn, M, P No

F14 French Canadian c.529G→A A177T 7 Horn, M, P No

F15 White British c.[IVS510+1 G→A]+[529G [Exon 6 splice lntron Het. M, P No

→Al donor 6: 7

site]+[A177T]

F16 German c.529G→A A177T 7 Horn , nps No

F17 White American c.[132T→A]+[655T→C] [C44X]+[Y219H] 2; 8 Het, M, P No

F18 White American c.[179T→G]+[529G→A] [L60R]+[A177T] 3; 7 Het, M, P No

F19 Italian c.[488C→T]+[529G→A] [T163I]+[A177T] 6; 7 Het, nps No

F20 Italian c.[488C→T]+[529G→A] [T163I]+[A177T] 6; 7 Het, M, P No

F21 Mixed white c.[529G→A]+[554T→G] [A177T]+[V185G] 7 Het, M, P No

Canadian/ Hungarian

Nucleotides are numbered from the A of the initiation codon (ATG) in the nucleotide sequence NM_024570. Abbreviations: fs, frameshift. Horn, homozygous in affected individual; Het, heterozygous in affected individual; M, mutation identified in mother; P, mutation identified in father; nps, no parental samples.

RNASEH2B (FLJ11712) is the ortholog of yeast Rnh2Bp FLJ11712 encodes a 308 amino-acid protein of previously undefined function. Semi-quantitative RT-PCR indicates that FLJ11712 is detectable in a wide range of human tissues, suggesting ubiquitous expression (data not shown).

A database search using PSI-BLAST (Altschuhl et al., Nucleic Acids Res 25:3389-3402 (1997)) revealed significant similarity (E = 4x10 ^"5 ) between human FLJ11712 and a Saccharomyces cerevisiae protein, Rnh2Bp (Rnh202), after 4 search iterations (Figure 8). Rnh2Bp is an essential component of the yeast RNase H2 enzyme complex (Jeong et al., Nucleic Acids Res 32:407-414 (2004)). This observation indicated that FLJ11712 might function in an equivalent complex in mammals, and raised the possibility that additional RNase H components might also be mutated in AGS.

RNASEH2C (AYP1) is the AGS3 gene

A SNP array genome-scan was performed on 6 consanguineous families,

5 Pakistani (F30-34), and 1 Bangladeshi (F35). This, along with subsequent microsatellite genotyping (Figure 9) identified a new locus, AGS3, on chromosome 11q13.2, with a maximum multipoint lodscore of 4.54 at 66.8cM (Marshfield Genetic Map), between markers D11S4205 and D11S913. Additionally, genotypes in the 5 Pakistani families suggested the presence of an ancestral haplotype (bold genotypes, Figure 9), which along with the small homozygous region in family F30 refined the AGS3 critical interval to a 4.9cm interval between D11S4205 and D11S987.

Notably, AYP1 , encoding a protein which biochemically copurifies with human RNASEH2A (Frank et al., Proc Natl Acad Sci USA 95:12872-

12877 (1988)), lies within this critical interval (Fig 3) AYP1 was therefore sequenced and nonconservative missense mutations were identified in these 6 families (see Table 3 and Figure 9). A homozygous mutation at codon 69 (R69W) was present in all affected individuals from the 5 Pakistani families that shared the apparent ancestral haplotype. A second different homozygous mutation, K143I, was present in the Bangledeshi family. The mutations segregated with the disease within the families, and neither mutation was detected in at least 172 Asian control alleles, RT- PCR expression profiling of AYP1 demonstrated a similar widespread expression pattern to FLJ 11712.

Table 3. AYP1 mutations identified in 6 Aicardi-Goutieres syndrome families

Family Ethnicity Nucleotide alterations Amino acid Exon Segregation Parental alterations consanguinity

AGS30 Pakistani c.205C→T R69W 2 Horn, M, P Yes

AGS31 Pakistani c.205C→T R69W 2 Horn, M, P Yes

AGS32 Pakistani c.205C→T R69W 2 Horn, M, P Yes

AGS33 Pakistani c.205C→T R69W 2 Horn, M, P Yes

AGS34 Pakistani c.205C→T R69W 2 Horn, M, P Yes

AGS35 Bangladeshi c.428A→T K143I 3 Horn, M, P Yes

Nucleotides are numbered from the start of the initiation codon (ATG) (for the protein transcript AF312034). Abbreviations: Horn, homozygous in affected individual; M, mutation identified in mother; P, mutation identified in father

Concurrently, the inventors established that AYP1 is the human ortholog of S. cerevisiae Rnh2Cp, a second subunit of yeast RNase H2 (Jeong et al., Nucleic Acids Res 32:4407-414 (2004)). A database search with a Kluyveromyces waltii open reading frame identifies both human AYP1 and S. cerevisiae Rnh2Cp within 4 search iterations (Figure 8).

RNASEH2A is the AGS4 gene

RNASEH2A is an 8 exon gene spanning 7 kbp of genomic sequence on chromosome 19p13.13, and encodes a 299 amino acid protein. RNASEH2A did not co-localise with any genetically mapped AGS loci. However, review of the SNP-array genome scan data identified a small region of homozygosity at the RNASEH2A locus in a previously described consanguineous AGS family of white Spanish ancestry (see Sanchis et al., J Pediatr 146:701-705, 2005) (Fig 4a). Sequencing in the 2 affected children from this family identified a single homozygous c.109G→A transversion, resulting in a G37S non-conservative missense mutation (Fig 4b). This amino acid is absolutely conserved from humans to archaebacteria (Fig 4c), and lies at a turn just at the end of the first β-sheet in the floor of the predicted substrate binding cleft, close to the RNase H catalytic site (Fig 4d) (Chapados et al., J MoI Biol 307:541-546, 2001).

The mutation was not detected in 178 Caucasian control alleles and both parents were heterozygous for the mutation.

RNASEH2B (FLJ11712), RNASEH2C (AYP1) and RNASEH2A form a complex in vitro, with RNase H2 activity

Sequence homology with the S. cerevisiae Rnh2Ap-Rnh2Bp-Rnh2Cp complex predicts that the RNASEH2B, RNASEH2C and RNASEH2A proteins will form a protein complex with RNase H2 activity (Fig 5a). To confirm this, the 3 genes were cloned using the Gateway system into epitope-tagged mammalian expression vectors (pCGT-Dest (T7) and

pCDNAS.I mychis-Dest) and all three constructs transiently co-transfected into HEK293 cells, lmmunoprecipitation with anti-myc antibodies against C-terminally tagged FLJ11712-myc, pulled down N-terminally T7-tagged AYP1 and T7-tagged RNASEH2A (Fig 5b) confirming that the three subunits interact in vitro.

A previously described fluorometric assay (Parniak et al., Anal Biochem 33-39, 2003) was adapted to test this complex for enzymatic activity. Fluorescein-labelled oligonucleotides were annealed to a complementary DABCYL-labelled DNA oligonucleotide that quenched fluorescence (Fig. 5c). Enzymatic cleavage of the fluorescein-oligonucleotide resulted in release of the fluorescein from the adjacent quencher molecule, generating a fluorescent signal (Fig 5c).

The inventors found that 'Oligo C, a RNA:DNA duplex, was efficiently cleaved by the immunoprecipitated complex, confirming that this complex exhibits RNase H activity (Fig 5d). Moreover, the immunoprecipitated complex possesses the requisite Type 2' RNase H activity, as it recognises a single ribonucleotide embedded in a DNA-DNA duplex, and efficiently cleaves 'Oligo B', in contrast to E.coli RNase H, a Type V RNase H. As expected, 'Oligo A' is not cleaved since it is synthesised using nuclease resistant 2-O'methylRNA chemistry.

Mutations in the RNase H2 complex reduce enzymatic activity Selected pathogenic mutations (FLJ11712, A177T, T1631 ; AYP1 R69W, K143I; RNASEH2A G37S) were introduced using site-directed mutagenesis of their respective genes, and expressed by transient transfection of HEK293 cells, lmmunoprecipitation was performed to assess their effect on complex stability, and enzymatic activity was assayed (Fig 6a and 6b). These experiments demonstrate that the

RNASEH2A G37S mutation has no effect on complex stability but, as expected for a mutation at the catalytic site, markedly reduces enzyme activity (Fig 6b). In the case of the AYP1 mutation, there is reduction in the amount of AYP subunit pulled down in complex with the other subunits. This is associated with a reduction in enzyme activity. This suggests that reduced enzyme activity is the result of the mutation disrupting complex formation.

Example 2: RNase H2 activity assay in poly (I:C) HCT 116 cells.

The inventors' observations are that innate immunity is inappropriately activated in AGS, suggesting that RNase H2 is involved in an innate immune response. To test this poly(l:C), a synthetic analogue of dsRNA was used, which is associated with viral infection and known to stimulate the innate immune response by inducing cytokine production via the dsRNA-response protein kinase (PKR) or the Toll-like receptor 3 (TLR3). Semi-confluent HCT116 colon cancer cells were treated over a time course with 25μg/ml poly(l:C) (Invivogen, USA) in RPMI1640 with 10%FCS and 1 % penicillin/streptomycin. Protein lysate was prepared from the cells and fluorometric RNase H2 assays were performed.

The results are shown in Figure 10. A two-fold increase in enzyme activity at 2 hours was observed, suggesting that RNase H2 lies downstream in this pathway and thus may be an effector of innate immune response.

Example 3: Polyclonal Antibody

Polyclonal antibodies were generated by expressing an RNase H2 immunogen as a bacterial expression protein (see "Antibodies: A Laboratory Manual" ed Harlow and Lane, Cold Spring Harbor Laboratory

Press (1 December 1988) ISBN 978-0879693145). The expression protein was purified using a GST tag on the B subunit. The GST tag was subsequently cleaved off using PreScission protease (Amersham) and the resultant protein injected into sheep (Eurogentec). Freund's adjuvant was used for the injections. A first injection was made using the complete adjuvant, followed by subsequent boost(s) with incomplete adjuvant. The resulting sheep sera has been immunoaffinity purified using columns containing the bound antigen complex. The antigen complex was the same as that used as the immunogen. Western blot (see Fig 11 ) shows that the antibody can detect overexpressed subunits of the correct molecular weight, ie. that the antibody has specificity against the RNase H2A/H2B and H2C subunits. An additional band ( ^* ) may represent one of the endogenous proteins. The identity of the bands in Western blot results was confirmed by reprobing with anti-tag antibody.

The immunogen sequences used were:

For anti- RNASEH2B (with residual linker from protease cleavage)

LEVLFQGPLGSPEFPSTSLYKKAGSTMAAGVDCGDGVGARQHVFLVSE YLKDASKKMKNGLMFVKLVNPCSGEGAIYLFNMCLQQLFEVKVFKEKHH SWFINQSVQSGGLLHFATPVDPLFLLLHYLIKADKEGKFQPLDQVWDNV FPNCILLLKLPGLEKLLHHVTEEKGNPEIDNKKYYKYSKEKTLKWLEKKVN QTVAALKTNNVNVSSRVQSTAFFSGDQASTDKEEDYIRYAHGLISDYIPK ELSDDLSKYLKPEPSASLPNPPSKKIKLSDEPVEAKEDYTKFNTKDLKTE KKNSKMTAAQKALAKVDKSGMKSIDTFFGVKNKKKIGKV (SEQ ID NO 52).

For anti-RNASEH2C

MESGDEAAIERHRVHLRSATLRDAVPATLHLLPCEVAVDGPAPVGRFFT PAIRQGPEGLEVSFRGRCLRGEEVAVPPGLVGYVMVTEEKKVSMGKPD

PLRDSGTDDQEEEPLERDFDRFIGATANFSRFTLWGLETIPGPDAKVRG ALTWPSLAAAIHAQVPED (SEQ ID NO 53).

For anti-RNASEH2A

MDLSELERDNTGRCRLSSPVPAVCRKEPCVLGVDEAGRGPVLGPMVYA

ICYCPLPRLADLEALKVADSKTLLESERERLFAKMEDTDFVGWALDVLSP

NLISTSMLGRVKYNLNSLSHDTATGLIQYALDQGVNVTQVFVDTVGMPE

TYQARLQQSFPGIEVTVKAKADALYPVVSAASICAKVARDQAVKKWQFV

EKLQDLDTDYGSGYPNDPKTKAWLKEHVEPVFGFPQFVRFSWRTAQTIL

EKEAEDVIWEDSASENQEGLRKITSYFLNEGSQARPRSSHRYFLERGLE

SATSL (SEQ ID No 54).

Example 4: Lymphoblastoid Cell Lines (LCL)

Lymphoblastoid cell lines (LCL) were generated by infecting primary B cells from AGS patients with EBV to create continuously proliferating cell lines. Suitable methodology for creating LCLs is well-know in the art, for example Penno et al., Methods in Cell Science 15(1 ):43-47, 1993. LCLs have been transformed for representative patient mutations in the subunits.

The following LCLs were created:

Table 4

Cellular enzyme activity can be assayed using the described fluorescent RNase H activity assay. As well as end point assays (Figure 13), kinetic assays can also be performed by timed measurements (eg every five minutes) on a Perkin Elmer Victor III that has a built-in temperature regulation (set at 37 ⁰ C). An exemplary activity assay on RNase H2 mutant lymphoblastoid cell lines is shown in Figure 13.

Example 5: Recombinant RNase H2

Recombinant, enzymatically active RNase H2 protein complex has been synthesised and purified (see Figure 15a) using a polycistronic bacterial expression construct containing the three genes. Activity of mutant and wild type complexes can be assessed by enzyme assays as shown in Figure 15b.

Example 6: RNase H2 to suppress viral replication

RNA-DNA intermediates are essential for viral replication. Therefore upregulation of endogenous RNase H2 activity is expected to be a useful host response to suppress viral replication. We have evidence to support this expectation, showing that there are significant increases in enzymic activity within several hours of treatment of HCT116 cells with dsRNA (polyl:C), a potent activator of innate immune signalling (see Figure 10).