PHAGE RESISTANT LACTIC ACID BACTERIA

Field of the invention

The present invention relates to a method for the construction of bacteriophage- insensitive Streptococcus thermophilus mutants, whereby a Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) system is attenuated, leading to phage resistance being conferred by an alternative mechanism.

Background of the invention

The thermophilic lactic acid bacterium Streptococcus thermophilus is widely used as a starter culture to improve the texture and flavour of many yoghurt and cheese products (Mora et al. (2002) Genetic diversity and technological properties of Streptococcus thermophilus strains isolated from dairy products. J Appl Microbiol, 93, 278-287). Consistent predation by (bacterio)phages, however, is still a major cause of economic losses in the dairy industry worldwide - despite a growing genetic and technological knowledge of both the hosts and phages (Goh, YJ et al (201 1 ) Specialized adaptation of a lactic acid bacterium to the milk environment: the comparative genomics of Streptococcus thermophilus LMD-9. Microb Cell Fact, 10 Suppl 1 , S22). To safeguard against losses of this type, it is necessary to develop robust and diverse Bacteriophage Insensitive Mutants (BIMs) of S. thermophilus starters for use in rotation.

Host defences against bacteriophage attack are frequently mediated by the Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) systems in S. thermophilus (Mills, S. et al. (2010). CRISPR analysis of bacteriophage-insensitive mutants (BIMs) of industrial Streptococcus thermophilus-implications for starter design. J Appl Microbiol, 108, 945-955), of which there are four known types in this bacterial species (Sinkunas, T et al. (2013). In vitro reconstitution of Cascade-mediated CRISPR immunity in Streptococcus thermophilus. EMBO J, 32, 385-394.). It is thought, however, that only CRISPR1 and CRISPR3 are active in S. thermophilus, being the only two systems which have been shown to provide acquired immunity against phages (Sinkunas, T. et al., 2013). CRISPR1 and 3 are both classified as Type ΙΓ systems, based on the structure and composition of the cas gene + repeat-spacer locus. The mechanism by which the CRISPR1 and CRISPR3 systems provide immunity from bacteriophage attack has been well characterised, and follows phage attachment and DNA injection into the host cell.

The non-native DNA triggers the CRISPR-associated (Cas) gene products to target and incorporate a short phage-derived nucleotide segment ("spacer") into the adjacent repeat-spacer locus on the host chromosome, a mechanism which is presumed to be mediated by the cas7 gene (Barrangou, R., et al. (2007). CRISPR provides acquired resistance against viruses in prokaryotes. Science, 315, 1709-1712). Subsequently, the system targets and degrades the incoming viral genome through the activity of the CRISPR-associated (Cas) 9 gene product (Sapranauskas, R., et al. (201 1 ). The Streptococcus thermophilus CRISPR/Cas system provides immunity in Escherichia coli. Nucleic acids research, 39, 9275-9282). For this reason, cas9 is now described as the 'signature' gene of the Type II CRISPR systems (Makarova, K. S., et al. (201 1 ). Evolution and classification of the CRISPR-Cas systems. Nature Reviews Microbiology, 9, 467-477) to which both CRISPR1 and 3 belong.

It is thought that, of the four systems, CRISPRI -mediated resistance is favoured by S. thermophilus, due to a higher frequency of spacer incorporation in this locus in some strains (Deveau, H., et al. (2008)). Phage response to CRISPR-encoded resistance in Streptococcus thermophilus. Journal of bacteriology, 190, 1390-1400; Mills et al (2010. Despite the effectiveness of the CRISPR1 system in providing acquired immunity to S. thermophilus against attacking phages, reliance on this natural mechanism in an industrial setting is not favourable. Deveau et al. (2008) showed that in order to evade the CRISPR1 system, the attacking phage merely needs to possess a single nucleotide polymorphism (SNP) in the segment from which the initial spacer was derived - highlighting the frequent instability of the system. The need for an alternative, more stable mechanism of phage-resistance (and a method to select for it) is evident. For these reasons, two of the cas genes (cas7 and cas9) forming part of the CRISPR1 - Cas system, were selected for targeted inactivation in this study.

The method of employing antisense RNA (previously referred to as micRNA) plasmid constructs to inhibit the expression of specific genes in bacteria and bacteriophages has been well established (Coleman, J. et al. (1985). A novel immune system against bacteriophage infection using complementary RNA (micRNA). Nature, 315, 601 -603; Kim & Batt, (1991 ). Antisense mRNA-Mediated Bacteriophage Resistance in Lactococcus lactis subsp. lactis. AppI Environ Microbiol, 57, 1 109-1 1 13. The L. lactis high copy plasmid pNZ44 was devised by McGrath, S. et al (2001 ) (Improvement and optimization of two engineered phage resistance mechanisms in Lactococcus lactis. AppI Environ Microbiol, 67, 608-616) to optimize an antisense strategy to silence phage genes involved in DNA replication, thereby increasing the phage resistance of the host. However, to our knowledge, the system has not been used to inactivate CRISPR-Cas genes prior to bacteriophage infection. The present invention provides a method to bias the generation of spontaneous BIMs in a manner which favours a CRISPR-independent mechanism of resistance. The invention also details a method whereby a phage- resistant mutant containing both CRISPR and non-CRISPR elements may be selected for.

Detailed description of the invention

In a first aspect, the invention provides a method for the construction of a bacteriophage insensitive mutant of a Streptococcus thermophilus parent strain comprising the steps of

a. inactivating the CRISPR resistance mechanism of the parent strain;

b. exposing the parent strain obtained in step a) to a bacteriophage;

c. isolating bacteriophage insensitive mutants;

d. optionally: comparing the CRISPR loci of the parent strain with the CRISPR loci of the bacteriophage insensitive mutants; and

e. optionally: selecting bacteriophage insensitive mutants of which the CRISPR loci is identical to the CRISPR loci of the parent strain.

The bacteriophage insensitive mutant shall be referred to in the remainder of the text as BIM.

The advantage of the method of the invention is that the BIM's have acquired their phage resistance due to a mechanism different from the CRISPR resistance mechanism. In order to avoid that in step b) of the method, CRISPR BIM's are generated, the CRISPR resistance mechanism must be inactivated in step a). Step a) - According to a preferred embodiment of the method of the invention, the CRISPR resistance mechanism is inactivated by introducing into the parent strain one or more DNA constructs comprising a promoter followed by one or more cas genes or a part thereof or a nucleotide sequence at least 90% identical to the cas gene in the reverse orientation such that the cas gene is transcribed into the corresponding antisense RNA which subsequently binds to the target cas mRNA thereby silencing the cas gene.

The promoter may be any suitable promoter such as pNZ44, derived from Lactococcus lactis (McGrath et al., 2001 ).

Suitable cas genes that may be used are cas genes associated with any functional CRISPR system may be used, provided that the function it performs is essential to the incorporation of additional spacers and/or the successful execution of acquired immunity against attacking phages. Preferred embodiments are cas7 and cas9, associated with the CRISPR1 system in S. thermophilus or cas5/csn1, casl or csn2 associated with CRISPR3. In the method of the invention a nucleotide sequence may be used that comprises only a part of the cas gene, for instance at least 100 base pairs, or at least 200 base pairs, or at least 300 base pairs. The advantage of this method is to increase the efficiency with which the antisense plasmids are constructed by using a shorter insert. However, preferably the full length cas gene is used for maximum efficiency of gene silencing after transcription to mRNA. Furthermore, in the method of the invention a nucleotide sequence may be used that is at least 90% identical, more preferably at least 91 %, more preferably at least 92%, more preferably at least 93%, more preferably at least 94%, more preferably at least 95%, more preferably at least 96%, more preferably at least 97%, more preferably at least 98%, more preferably at least 99% and most preferably 100% identical to the nucleotide sequence the cas gene or the part thereof that is used.

In a preferred embodiment, the CRISPR resistance mechanism is inactivated by introducing into the parent strain one or more DNA constructs comprising a promoter followed by two or more cas genes or a part thereof.

Two or more DNA-constructs may also be used in step a) whereby each DNA- construct may comprise the same or a different cas gene. This method may provide a significant advantage by inactivating two or more functional CRISPR-Cas systems simultaneously, in order to rule out potential transference of activity from an inactivated CRISPR1 system to CRISPR3 or 4, which is equally undesirable for the reasons stated above.

According to a another preferred embodiment of the method of the invention, the CRISPR resistance mechanism is inactivated by introducing into the parent strain one or more DNA constructs comprising a catalytically inactive Cas9 protein and one or more single guide RNAs (sgRNAs) for transcriptional repression of the one or more cas gene(s) in the Streptococcus thermophilus parent strain. The gene encoding the catalytically inactive Cas9 protein and a customizable single guide RNA (sgRNA) are co- expressed and the Cas9-sgRNA complex may bind to DNA elements complementary to the sgRNA and may cause a steric block that halts transcript elongation by RNA polymerase, resulting in the repression of the cas9 gene. This method has been shown to be capable of repressing target genes in for instance Escherichia coli (Qi .S. et al. (2013) Repurposing CRISPR as an RNA-guided platform for sequence-specific control of gene expression Cell. Feb 28;152(5):1 173-83). This system, which is called CRISPR interference (CRISPRi) by Qi L.S. et al., can efficiently repress expression of targeted genes in Escherichia coli, with no detectable off-target effects. CRISPRi can be used to repress multiple target genes simultaneously, and its effects are reversible.

Step b) - Exposing the parent strain obtained in step a) to a bacteriophage may be carried out in any suitable medium, for instance in an aqueous solution such as a buffered aqueous solution or in a soft agar medium or in milk. In a preferred embodiment, exposing the parent strain to a bacteriophage is carried out in a soft agar medium. In another preferred embodiment, exposing the parent strain to a bacteriophage is carried out in milk. The milk may be incubated overnight or until clotting is observed. The parent strain used in the method of the invention may be pre-treated in order to increase the genetic diversity and to increase the number of the BIMs. This pre- treatment may be carried out by methods known in the art, such as chemical mutagenesis or by irradiation with UV-light. The - optionally pre-treated - parent strain may be exposed to one type of bacteriophage or to multiple different bacteriophages, for instance to 2, 3, 4 or 5 different type of bacteriophages.

Step (c) -The suspension or the incubated (clotted) milk obtained in step (a) of the method of the invention may be plated on agar plates. After incubating the agar plates at a temperature at which Streptococcus thermophilus may grow, colonies may appear which represent the BIMs. The colonies may be purified to obtain a single strain BIM according to methods known in the art.

Step (d) - In step (d) of the method of the invention, the CRISPR loci of the BIMs obtained in step (c) of the method of the invention are analyzed for their length (in base pairs) and/or sequenced and compared with the CRISPR loci of the bacteriophage sensitive parent strain.

Step (e) - In step (e) of the method of the invention only those BIMs are selected of which the CRISPR loci are identical to the CRISPR loci of the parent strain. The advantage of the method of the invention is, that the selected BIMs have acquired a phage resistance mechanism that is different from CRISPR and therefore based on an alternative phage resistance mechanism. As a result, the BIMs obtained by the method of the invention have a more stable phage resistance compared to a CRISPR BIM of which it is known that phages can rapidly evolve to overcome these spacer additions through single nucleotide alterations in the appropriate genomic region.

In another embodiment of the method of the invention, the one or more DNA constructs that were introduced in step a) of the method of the invention may be removed after step e) The advantage of this step is that the CRISPR defence mechanism is once again active against the bacteriophages.

The Streptococcus thermophilus parent strain used in the method of the invention, may be either be sensitive or be insensitive to the bacteriophage, such as insensitive due to the previous addition of a CRISPR spacer. In the case of a sensitive parent strain, inactivation of the CRISPR defence mechanism prevents the formation of CRISPR BIMs in step b) of the method of the invention. In the case of the insensitive parent strain (such as due to the previous addition of a CRISPR spacer), the parent strain is first made sensitive by inactivation of the CRISPR defence mechanism which then also prevents the formation of phages that have overcome the CRISPR-mediated defence mechanism in step b) of the method of the invention. This may lead to a BIM which is resistant to one or more phages by both CRISPR and non-CRISPR mechanisms.

In a second aspect, the invention provides a bacteriophage-insensitive mutant of a Streptococcus thermophilus parent strain wherein the CRISPR loci of the bacteriophage insensitive mutant are identical to the CRISPR loci of the parent Streptococcus thermophilus strain and which are obtainable by the method of the first aspect of the invention.

Alternatively, the present invention relates to a bacteriophage-insensitive mutant of a Streptococcus thermophilus parent strain wherein the CRISPR loci of the bacteriophage-insensitive mutant are not identical to the CRISPR loci of the parent Streptococcus thermophilus strain and which bacteriophage-insensitive mutant further comprises a non CRISR mediated bacteriophage resistance mechanism. Preferably, the non CRISPR mediated bacteriophage resistance mechanism is not a restriction- modification system. Preferably, the non CRISPR mediated bacteriophage resistance mechanism is a resistance mechanism derived from Streptococcus thermophilus, or preferably derived from the Streptococcus thermophilus parent strain. The advantage of using both a CRISPR and a non CRISPR resistance mechanisms in a BIM is an improved phage resistance. So called 'double hurdle' BIMs have a broader phage resistance, i.e. a defence against several different bacteriophages. Furthermore, double hurdle BIMs have an increased phage robustness, showing resistance for several rounds of exposure to a bacteriophage.

In a preferred embodiment, the present bacteriophage insensitive mutant has an increased sedimentation rate (or an increased pellet weight) and/or an increased chain formation compared to the Streptococcus thermophilus parent strain. Preferably, the present bacteriophage insensitive mutant has a pellet weight increase of least 10% of the pellet weight of the parent strain, more preferably at least 20%, more preferably at least 50%, more preferably at least 100%, preferably at least 150%, more preferably at least 200%, more preferably at least 250%, even more preferably at least 300% and most preferably at least 350% of the pellet weight of the parent strain. Preferably, the present bacteriophage insensitive mutant has a percentage increase of average chain length or average cells per chain (CPC) of at least 50%, more preferably at least 100%, preferably at least 150%, more preferably at least 200%, more preferably at least 250%, even more preferably at least 300% and most preferably at least 350% of the averaged CPC of the parent strain.

In a third aspect, the invention provides a process for the production of a dairy product such as a fermented milk product or cheese comprising the use of one or more of the bacteriophage insensitive mutant of a bacteriophage-sensitive Streptococcus thermophilus parent strain as disclosed hereinbefore.

In a fourth aspect, the invention provides the use of the bacteriophage-insensitive mutant derived from a bacteriophage-sensitive Streptococcus thermophilus parent strain as disclosed hereinbefore in a process for the production of a dairy product, such as a fermented milk or cheese. Figure legends

Figure 1 - Plasmid preparations of S. thermophilus strain 100-E and derivatives. Lanes 1 , 10: 1 kb Full Scale DNA Ladder (Fisher Scientific, U.S.A.), L2: 100-E parent, L3: BIM100-E-D1A-L7, L4: BIM100-E-D1A-L7::pNZ44, L5: BIM100-E-D1A-L7::pNZ44+100- ECas7i, L6: BIM100-E-D1A-L7::pNZ44+100-ECas9i, L7: BIM100-E-D1A- L7::pNZ44+100-ECas9i (cured), L8: BIM100-E-D1A-L5, L9: BIM100-E-D1A- L5::pNZ44+100-ECas9i.

Figure 2 - PCR amplification of pNZ44 MCS regions. Lanes 1 , 12: 1 kb Full Scale DNA Ladder , L2: pNZ44 plasmid DNA (+ve control; indicated by arrow), L3: 100-E parent, L4: BIM100-E-D1A-L7, L5: BIM100-E-D1A-L7::pNZ44, L6: BIM100-E-D1A-L7::pNZ44+100- ECas7i, L7: BIM100-E-D1A-L7::pNZ44+100-ECas9i, L8: BIM100-E-D1A- L7::pNZ44+100-ECas9i (cured), L9: BIM100-E-D1A-L5, L10: BIM100-E-D1A- L5::pNZ44+100-ECas9i, L1 1 : Negative control.

Figure 3 - GTG RAPD fingerprinting of S. thermophilus parent strains, BIMs and transformants used in this study. Lane 1 : 1 kb Full Scale DNA Ladder (, L2: 100-E parent, L3: 100-E::pNZ44+100-ECas7i, L4: 100-E::pNZ44+100-ECas9i, L5: BIM100-E-D1A-L7, L6: BIM100-E-D1A-L7::pNZ44, L7: BIM100-E-D1A-L7::pNZ44+100-ECas7i, L8: BIM100- E-D1A-L7::pNZ44+100-ECas9i, L9: BIM100-E-D1A-L7::pNZ44+100-ECas9i (cured), L10: BIM100-E-D1A-L5, L1 1 : BIM100-E-D1A-L5::pNZ44+100-ECas9i, L12: Negative control.

Figure 4 - PCR amplification of CRISPR (C1 , C2, C3) loci of S. thermophilus strain 100- E and its derivatives. Lanel : 1 kb Full Scale DNA Ladder, L2: 100-E parent C1 , L3: BIM100-E-D1A-L7::pNZ44+100-ECas9i C1 , Lanes 4, 7 & 10: -ve controls, L5: 100-E parent C2, L6: BIM100-E-D1A-L7::pNZ44+100-ECas9i C2, L8: 100-E parent C3, L9: BIM100-E-D1A-L7::pNZ44+100-ECas9i C3. Figure 5 - Observed sedimentation of S. thermophilus strain 100-E parent (a) and its derived BIMs. (b) BIM100-E-D1A-L-7 (CRISPR BIM), (c) BIM100-E-D1A-L-5 (non- CRISPR BIM). Figure 6 - Representative images of S. thermophilus 100-E (and derived BIMs) cell chains visualised using a confocal laser scanning microscopy, (a) 100-E WT, (b) BIM100-E-D1A-L7, (c) BIM100-E-D1A-L5, (d) BIM100-E-DH51 , (e) BIM100-E-DH61 . Figure 7 - PCR amplification of pNZ44 MCS regions. Lane 1 : 1 kb Full Scale DNA Ladder, L2: pNZ44 plasmid DNA (positive control), L3: negative control, L4: 100-E WT, L5: BIM100-E-D1A-L7, L6: BIM100-E-D1A-L7::pNZ44+100-ECas7i, L7: BIM100-E- DH51 , L8: BIM100-E-DH61. Figure 8 - Observed sedimentation profile of S. thermophilus 100-E and derived BIMs. (a) 100-E WT, (b) BIM100-E-D1A-L7, (c) BIM100-E-DH51 , (d) BIM100-E-DH61.

Figure 9 - GTG RAPD fingerprinting of S. thermophilus 100-E parent strain and derived BIMs. Lane 1 ; 8: 1 kb Full Scale DNA Ladder, L2: 100-E WT, L3: BI M100-E-D1A-L5, L4: BIM100-E-D1A-L7, L5: BIM100-E-DH51 , L6: BIM100-E-DH61 , L7: negative control.

Figure 10 - GTG RAPD fingerprinting of S. thermophilus 100-F parent strain and transformants. Lane 1 : 1 kb Full Scale DNA Ladder, L2: 100-F WT, L3: BIM100-F- NGBD1A-L2, L4: BIM100-F-NGBD1A-L2::pNZ44, L5: BIM100-F-NGBD1A- L2::pNZ44+100-FCas9i, L6: BIM100-F-NGBD1A-L2::pNZ44+100-FCas9i (cured), L7: negative control.

Figure 1 1 - PCR amplification of pNZ44 MCS regions. Lanes 1 : 1 kb Full Scale DNA Ladder, L2: pNZ44 plasmid DNA (positive control) , L3: negative control, L4: 100-F WT, L5: BIM100-F-NGBD1A-L2, L6: BIM100-F-NGBD1A-L2::pNZ44, L7: BIM100-F- NGBD1 A-L2::pNZ44+100-FCas9i, L8: BIM100-F-NGBD1 A-L2::pNZ44+100-FCas9i (cured).

Figure 12 - GTG RAPD fingerprinting of S. thermophilus 100-F parent strain and transformants. Lane 1 ; 8: 1 kb Full Scale DNA Ladder, L2: 100-F WT, L3: 100- FWT::pNZ44 (plasmid control), L4: 100-F::pNZ44+100-FCas7i, L5: 100-F::pNZ44+100- F2Casi, L6: 100-F::pNZ44+100-F2CasSWi, L7: negative control. Figure 13 - PCR amplification of pNZ44 MCS regions. Lane 1 : 1 kb Full Scale DNA Ladder, L2: 100-F WT, L3: negative control, L4: pNZ44 plasmid DNA (positiveve control), L5: 100-FWT::pNZ44 (plasmid control), L6: 100-F::pNZ44+100-FCas7i, L7: 100-F::pNZ44+100-F2Casi, L8: 100-F::pNZ44+100-F2CasSWi.

MATERIALS AND METHODS

1. Bacterial growth conditions

Individual strains of S. thermophilus were routinely grown from 20% Reconstituted Skimmed Milk (RSM) stocks, 20% glycerol stocks (Sigma Aldrich, Germany) or from a single colony overnight (ON) at 42°C in M17 Broth (Oxoid, U.K.) supplemented with 0.5% lactose (LM17) or on plates containing 10 g/L technical agar (Merck, Germany). In phage enumeration assays, as adapted from Lillehaug, D. ((1997) An improved plaque assay for poor plaque-producing temperate lactococcal bacteriophages. J Appl Microbiol, 83, 85-90). LM17 broth was supplemented with 0.25% glycine (Oxoid, U.K.), 10 mM CaCI ₂ (Oxoid, U.K.) and either 10 g/L (solid agar base) or 4 g/L (semi-solid overlay) technical agar. The semi-solid agar was sterilised by autoclaving at 121 °C for 15 minutes or by boiling for 7 minutes in a microwave whereas the solid agar was boiled for 7 minutes in a microwave. L. lactis NZ9000 was maintained as above with the following modifications: overnight (ON) cultures were grown at 30°C with the substitution of glucose (Sigma-Aldrich, Germany) in place of lactose. All transformants were maintained as above with the addition of chloramphenicol (Sigma- Aldrich, Germany) to a final concentration of 5 μg/ml (L/GM17+Cm5). 2. Isolation and selection of bacteriophages

Whey samples from dairy plants producing fermented milk products were obtained and analysed for the presence of phages against S. thermophilus 100-E using the spot assay described below under "Bacteriophage assays". Defined single plaques were isolated by twice single plaque purification on semi-solid overlays. Phages were then propagated as follows: 10 ml LM17 broth was inoculated (1 %) with a fresh ON culture of the appropriate host strain and incubated at 42°C for 1.0-2.5 hours. Then, a single plaque was added to the growing culture, mixed well and incubated for a further 2- 4 hours. The lysed culture was centrifuged and the supernatant filtered (0.45 μηη). The filtered supernatant was used as the phage stock for subsequent assays. Table 1 summarizes the phages that were obtained in this manner.

Table 1. A list of strains and phages used in this study

Strain Description Source

Streptococcus thermophilus 100-E Parent strain DSM, The Netherlands

L. lactis NZ9000 Transformation vector a)

BIM100-E-D1A-L7 CRISPR-mediated BIM UCC, Cork, Ireland

BIM100-E-D1A-L5 Non-CRISPR BIM

BIM100-E-DH51 Double Hurdle BIM

BIM100-E-DH61 Double Hurdle BIM

Streptococcus thermophilus 100-F Parent strain DSM, The Netherlands

BIM100-F-NGBD1A-L2 CRISPR-mediated BIM UCC, Cork, Ireland

Phage Description Source

Virulent phage of S.

cp100-E-D1A-L DSM, The Netherlands thermophilus 100-E

cp100-E-D2A-L

cp100-E-D3A-L

cp100-E-D4A-L

Virulent phage of S.

cpSTV84-D1A-L

thermophilus 100-F

cpSTV88-D1A-L cpNGB-D1A-L

a) Fernandez, L, et al (2000). Cloning, characterization, controlled overexpression, and inactivation of the major tributyrin esterase gene of Lactococcus lactis. AppI Environ Microbiol, 66, 1360-1368.

Streptococcus thermophilus 100-E = DS64900 was deposited on 15 July 2014 with the Centraal Bureau for Schimmelcultures, Uppsalalaan 8, 3508 AD in Utrecht, The Netherlands and received deposition number CBS138555. Streptococcus thermophilus 100-F = DS64985 was deposited on 6 May 2015 with the Centraal Bureau for Schimmelcultures, Uppsalalaan 8, 3508 AD in Utrecht, The Netherlands and received deposition number CBS139996.

3. BIMs of S. thermophilus 100-E and 100-F

BIMs against phage cp100-E-D1A-L were generated by adding 500 μΙ fresh ON culture of S. thermophilus 100-E and 10 to 100 μΙ undiluted phage lysate as indicated (phages isolated from a single plaque, titre approx. 1 x 10 ⁷ pfu/ml) to 4 ml soft LM17 agar, followed by spreading this suspension on solid agar. In the case of 100-F and where insufficient BIMs were isolated using ON incubation, the amount of culture added was increased to 1000 μΙ and the amount of phage lysate to 100 μΙ. Colonies, representing potential BIMs, growing in the top layer were twice single colony-purified and subjected to phage assays and CRISPR sequencing as described below. For 100-E, two BIMs were thus obtained and selected for characterization (see below): S. thermophilus BIM100-E-D1A-L5 and BIM100-E-D1A-L7. A further two BIMs, derived from BIM100-E-D1A-L7, were also generated in this manner (BIM100-E-DH51 and BIM100-E-DH61 ; Table 1 ). All 100-F BIMs (Table 1 ) were generated in this manner, with the modifications indicated above. Where insufficient numbers of BIMs were generated using this method, the procedure was repeated using the appropriate WT or transformed strain until an acceptable number of BIMs was isolated.

4. Bacteriophage assays

Spot assays were performed by seeding the LM17 semi-solid agar overlay with 500 μΙ fresh ON culture and applying 10 μ I of phage lysate in a grid format, as described by Dupont, K., et al, J. (2005) (Detection of lactococcal 936-species bacteriophages in whey by magnetic capture hybridization PCR targeting a variable region of receptor- binding protein genes. J AppI Microbiol, 98, 1001 -1009). Plates were then allowed to dry and incubated anaerobically ON at 42°C. A clear zone was assumed to indicate phage- mediated lysis of the bacterial lawn by the applied phage and was recorded as '+', whereas absence of lysis was recorded as

For phage enumeration, plaque assays were performed by adding 500 μΙ culture and 10 μ I of neat or appropriately diluted phage lysate to 4 ml semi-solid agar, followed by plating on LM17 agar plates as described above with subsequent ON incubation at 42°C. Efficiency of plaquing (EOP) was calculated by dividing the obtained titre of a given phage on the test strain by the titre of the same phage on the parent strain. 5. PCR screening & CRISPR locus sequencing

All BIMs generated were subjected to PCR profiling to confirm their relatedness to the relevant parent strain from which they were derived, either by RAPD PCR or by CRISPR locus sequencing. RAPD PCR was performed on single colonies of each parent strain and BIM as template for the reaction and using the '(GTG)5' RAPD profiling primer (Gevers, D., et al. (2001 ). Applicability of rep-PCR fingerprinting for identification of Lactobacillus species. FEMS Microbiol Lett, 205, 31 -36). The PCR conditions were as follows: 95°C x 10 min, followed by 30 cycles of 95 °C x 15 s, 40 °C x 30 s and 72 °C for 8 min with a final extension step of 72 °C for 16 min.

BIMs generated were purified and the CRISPR loci amplified by PCR and sequenced to determine acquisitions or alterations to the spacer content of the BIMs. CRISPR-1 , CRISPR-2 and CRISPR-3 repeat/spacer arrays for each strain were amplified individually using a single colony of the appropriate strain as template material for the PCR and primers described previously by Horvath, P et al. (2008). Diversity, activity and evolution of CRISPR loci in Streptococcus thermophilus. Journal of bacteriology, 190, 1401 -1412). Primers targeting the CRISPR4 locus of 100-F were designed specifically for this strain, as no homolog of CRISPR4 is present in 100-E. The PCR conditions were as follows: 95°C x 10 min, followed by 30 cycles of 95°C x 15 s, 55°C x 15 s and 72°C for either 2 min 45 s (CRISPR-1 ) or 1 min 10 s (CRISPR-2, CRISPR-3 and CRISPR-4) with a final extension step of 72 °C for 10 min.

The PCR generated products were visualised on a 1 % agarose (Fisher Scientific, USA) gel and purified using a PCR purification spin kit (Genomed, Germany). Sanger sequencing (of all PCR products and plasmids) was performed by MWG Biotech to verify the integrity of all plasmid constructs and to compare the sequences of the CRISPR loci of the BIMs to those of the corresponding parent strain (Eurofins, Germany). For CRISPR loci, this was performed by primer walking using synthetic primers based on a unique spacer of each repeat/spacer array in the internal regions of the sequences of the CRISPR loci, where required. CRISPRs were assembled using the Seqman program (DNAstar) and CRISPR arrays were visualised using the online CRISPR finder program (http://crispr.u-psud.fr). Table 2. PCR primers used in this study

a) Horvath, et al., 2008

b) This study

6. Preparation of competent cells

Competent cells were prepared as described by Holo & Nes ((1989). High- Frequency Transformation, by Electroporation, of Lactococcus lactis subsp. cremoris Grown with Glycine in Osmotically Stabilized Media. Appl Environ Microbiol, 55, 31 19- 3123), with the following modifications: A series of tubes containing 10 ml LM17 or GM17 (for L. lactis) broth and varying (from 0.2% to 2.4%) concentrations of either glycine (Sigma-Aldrich, Germany) or threonine (Sigma-Aldrich, Germany) were prepared and inoculated (1 %) with a fresh ON culture. The tubes were incubated at 42°C ON and examined for growth. LM17 broth containing 0.5% sucrose (Sigma-Aldrich, Germany; SLM17) supplemented with the highest level of glycine or threonine tolerated by the strains was used to prepare competent cells. Subsequently, 50 ml SLM17 or SGM17 broth supplemented with 1 % glycine (NZ9000) 2% (100-E) or 0.8% (100-F) threonine were inoculated (2%) with the appropriate strain and incubated until an OD ₆ oonm of approximately 0.5 was reached. In the case of 100-E, ampicillin (Sigma-Aldrich, Germany) was added to a final concentration of 20 μg ml, and incubation continued for a further 1 hour. All steps from this point onward were performed at either 4°C or on ice. The cells were centrifuged at 4500 rpm for 15 minutes to pellet and the supernatant discarded. The cells were then washed twice in 15 ml ice cold 0.5 M sucrose/10% glycerol (Sigma-Aldrich, Germany; SG) solution. Finally, the cells were resuspended in 1 ml SG and 100 μΙ aliquots were immediately stored at -80°C until electrotransformation as described below. 7. Construction of antisense plasmid vectors

The PCR primers used to amplify cas7, csn2 or cas9 from either S. thermophilus 100-E or 100-F are listed in Table 2, and were designed to incorporate the entire relevant gene including the Shine-Dalgarno (SD) sequence. To ensure that an antisense product was produced, all cas genes were cloned into the pNZ44 plasmid (containing a chloramphenicol resistance gene marker) in the reverse orientation relative to the p44 promoter, as described by McGrath, et al. (2001 ). In the case of pNZ44+100-F2Casi, the csn2 gene was cloned directly behind the p44 promoter, followed by cas7, whereas in the pNZ44+100-F2CasSWi construct, cas7 was cloned directly behind the p44 promoter followed by csn2. Firstly, the appropriate cas gene product was PCR amplified from S. thermophilus 100-E or 100-F. The PCR conditions were as follows: 95°C x 10 min, followed by 30 cycles of 95°C x 15 s, 55°C x 15 s and 72°C for either 3 min 15 s {cas9) or 1 min 10 s (cas7, csn2) with a final extension step of 72 °C for 10 min. The products were then purified using the JetQuick PCR product purification kit (Genomed, Germany). pNZ44 plasmid DNA was isolated from Lactococcus lactis NZ9000::pNZ44 using the GeneJet plasmid Miniprep kit (Thermo Scientific, U.S.A.). PCR products and plasmid were restricted separately ON at room temperature (RT) using the restriction enzymes (Roche, Germany) Pstl and either Ncol (100-E-derived cas7) or Xbal (100-E-derived cas9). The reaction mixture was as follows: 50 μΙ plasmid/product, 2 μΙ each restriction enzyme, 20 μΙ buffer H (Roche, Germany), 126 μΙ H ₂ 0). Restricted products were subjected to electrophoretic examination to check their integrity and to estimate relative amounts (for ligation, an approximate insert to vector ratio of 4:1 was considered ideal). Ligation of insert and vector was performed by incubating 1 μΙ T4 DNA ligase (Roche, Germany), 1 μΙ ligase buffer (Roche, Germany), 6 μΙ insert and 2 μΙ vector ON at RT. For constructs pNZ44+100-F2Casi and pNZ44+100-F2CasSWi, 3 μΙ of each appropriate insert was used in the ligation reaction. Antisense constructs used in this study are listed in Table 3.

Tab e 3. Antisense constructs used in this study

8. Electrotransformation & transformant selection

Prior to transformation, constructs were dialysed using 0.025 μηη MF membrane filters (Merck Millipore, Germany) for 10 mins against sdH ₂ 0. All constructs were generated in L. lactis NZ9000 prior to their subsequent transfer to S. thermophilus 100-E or 100-F. Electrotransformation was performed using freshly prepared competent cells as described above, with the following modifications: competent cells were defrosted on ice for 5 mins, 10 μΙ construct was added and the solution was gently mixed. The mixture was transferred to a pre-chilled 2 mm electroporation cuvette (Cell Projects, U.K.) and electroporated under the following conditions: 1 .75 kV (S. thermophilus) or 2.0 kV (L lactis)/200 ohm/25 F. 950 μΙ recovery broth (LM17 or GM17 with the addition of 20 mM MgCI ₂ and 2 mM CaCI ₂ (Sigma-Aldrich, Germany)) was immediately added and the transformed cells were recovered at 30°C (L. lactis) or 42°C (S. thermophilus) for 2.5 hrs. 100 μΙ of undiluted and, where appropriate, diluted transformed cells were plated on antibiotic selection plates (LM17 or GM17 containing 5 μg ml chloramphenicol (Sigma- Aldrich, Germany); L/GM17+Cm5) and incubated ON at the appropriate temperature. Colonies representing potential transformants were screened by PCR using a forward primer designed upstream of the multiple cloning site (MCS) of plasmid pNZ44 and the appropriate 'forward' insert primer (Table 2), to reduce the incidence of false positive detection. Presumed transformants were purified on LM17+Cm5 agar plates and subjected to CRISPR sequencing, phage sensitivity assays (as described above) and plasmid sequencing, which was performed using primers designed outside the MCS of pNZ44 (pNZ44fwd and rev; Table 2).

9. Plasmid curing

In order to cure introduced pNZ44 and pNZ44-derivative vectors, selected transformants were subjected to four ON passages at 42°C in LM17 without the addition of chloramphenicol. Overnight cultures were then ten-fold serially diluted in ¼ strength Ringers solution (Merck, Germany), or a plate streak performed, and individual colonies were assessed for sensitivity to chloramphenicol by streaking on LM17+Cm5 agar plates. Streaked colonies which showed no growth on LM17+Cm5 but growth on LM17 were defined as presumptive cured transformants and subjected to validation by CRISPR PCR and plasmid preparation as described above.

10. Sedimentation assays

S. thermophilus strains were routinely grown from 10% glycerol stocks, 20% Reconstituted Skimmed Milk (RSM) stocks or from single colonies overnight at 42°C in LM17 broth (as described in section 1 of the MATERIALS AND METHODS). The parent strains and BIMs were treated identically and after overnight incubation at 42°C, visual assessment of the cultures was performed to observe the growth characteristic of the cultures in broth. Only if the cultures were consistently observed to sediment to the base of the tube or along the wall of the tube was the phenotype considered relevant. In all cases, the parent strain was observed to sediment to a markedly reduced degree after overnight growth. For the optical density assay, the OD600 _nm of 1 ml of each culture (taken from the top of the test tube) after ON incubation was measured using a DU 730 spectrophotometer (Beckman Coulter, U.S.A.). 11. Microscopic assays

Morphological assessment and comparison of the parent strains and derived BIMs was performed via wet mount. 5 μΙ of fresh overnight culture was placed on a glass slide (in duplicate) and a cover slip immediately placed on top of the sample. Each sample was then visualised under 63X magnification using a confocal laser scanning microscope and a Zeiss LSM 5 Exciter (Carl Zeiss, Jena, Germany; excitation 488 nm). The percentage increase in chain length or cells per chain (CPC) of derived BIMs relative to the parent strains was calculated firstly by determining the average number of individual cells per chain in all samples by counting at least 20 chains per strain. The average increase in length was then expressed as a percentage using the following formula: (CPC _mu tant - CPCparent) CPCparent x 100%). In this and all cases, the unpaired student t-test was used to determine significant differences between the parent and derived BIMs datasets. 12. Heap Lawrence assay

The phage resistance robustness was assessed using a so called "Heap- Lawrence" assay as follows. An overnight (ON) culture of the strains to be assessed (wild type plus BIMs) was made by inoculating 250 μί of 10% RSM with 50uL stock culture and incubated O/N at 42°C. The next day, the ON culture was diluted 5 times with fresh 10% RSM, 50 μΙ diluted culture was added to an MTP containing 650μΙ 10% RSM. Cells were allowed to grow for 1 hour and then a high titer phage lysate was added. Strain and phage were incubated at 42°C. Acidification was monitored after 6-8 hrs after which strains and phage were left ON. The next day MTPs were centrifuged at 4000rpm for 10 minutes and supernatant, containing the phages, was transferred to a greiner tube. Part of the supernatant was mixed 1 : 1 with phage lysate in a new tube. This phage-mix was then used as a source of phage in another round of Heap-Lawrence as described for day 2. The procedure was repeated for many cycles, thereby allowing the phages to adept to overcome the strain's phage resistance. By monitoring the number of cycles it takes for a phage to become virulent (indicated by the inability of the strains to acidify the milk) an indication of the phage robustness is obtained. EXAMPLES

Example 1 CRISPR silencing of Parent and Bacteriophage Insensitive Mutants (BIMs) of S.

thermophilus 100-E

1. Cas9 silencing in S. thermophilus 100-E BIMs

1.1. Generation and analysis of 100-E CRISPR and non-CRISPR BIMs

Bacteriophages against S. thermophilus 100-E (4 in total) and BI Ms against these phages were isolated as described in the MATERIALS AND METHODS. Two BI Ms, S. thermophilus BI M100-E-D1 A-L5 and BI M100-E-D1 A-L7, were selected for further analysis by plaque assay and CRISPR sequencing, the results of which are shown in Tables 4 and 5, respectively.

Table 4. Relative efficiencies of plaquing (EOP) of phages of Streptococcus thermophilus strain 100-E and derived BIMs.

BIM was challenged with a phage lysate containing 1 x 10 pfu/ml phages.

Note 2: In derived BI M nomenclature, D1A-L denotes the phage against which the BI M was generated.

Table 5: Summary of CRISPR repeat/spacer loci content in S. thermophilus strain

The above results show that the mechanisms responsible for cp100-E-D1A-L resistance in BIM100-E-D1A-L5 and BIM100-E-D1A-L7 are different.

BIM100-E-D1A-L5 can be defined as a 'non-CRISPR BIM' due to the fact that CRISPR repeat/spacer locus is 100% identical to the parent strain in the three loci sequenced (table 5), combined with the apparent absence of any additional CRISPR- Cas loci on the whole sequenced genome. In addition, the sedimentation profile (Figure 5; Table 1 1 ), typified here (and in all cases) by cell aggregation at the bottom of the ON culture tube, and relative increase of cell chain length (Figure 6; Table 10) indicate that this BIM has undergone a mutation in a phage receptor that is a component of the cell envelope.

In the case of BIM100-E-D1A-L7, it was concluded that insensitivity to φ100-Ε- D1A-L was conferred by the CRISPR mechanism, due to the addition of one spacer at the leader end of the CRISPR1 repeat/spacer locus (Table 5). This 30 bp spacer showed 100% homology to a region spanning the start codon of a gene coding for a RecT recombinase protein on the genome of phage cp100-E-D1A-L, which was used in the challenge. This exact 30 bp segment was not found in the genomes of any of the other three phages infecting S. thermophilus 100-E (data not shown). This conclusion is supported by the complete absence in BIM100-E-D1A-L7 of the distinctive sedimentation profile (Figure 5; Table 1 1 ), and cell chain length increase (Figure 6, Table 10), indicative of a mutation in a component of the cell envelope (as seen in BIM100-E-D1A-L5 and double hurdle BIMs below).

1.2. Transformation, curing & phage sensitivities of BIM100-E-D1 A-L5 and BIM100- E-D1A-L7

S. thermophilus BIM100-E-D1A-L5 and BIM100-E-D1A-L7 were transformed using freshly isolated pNZ44+100-ECas9i and pNZ44, pNZ44+100-ECas7i and pNZ44+100-ECas9i, respectively, and potential transformants were screened using the confirmatory PCR as described in the MATERIALS AND METHODS. Four presumptive transformants, BIM100-E-D1A-L5::pNZ44+100-ECas9i (non-CRISPR control), BIM100- E-D1A-L7::pNZ44 (plasmid control), BIM100-E-D1A-L7::pNZ44+100-ECas7i (for use in 'double hurdle' BIM generation, see section 1.3) and BIM100-E-D1A-L7::pNZ44+100- ECas9i (proof of principle) were selected for validation and further analysis.

BIM100-E-D1A-L5::pNZ44+100-ECas9i and BIM100-E-D1A-L7::pNZ44+100- ECas9i were subjected to plasmid preparation (Figure 1 ), pNZ44 MCS PCR (Figure 2), GTG RAPD typing PCR (Figure 3) and CRISPR PCR (Figure 4) followed by sequencing, as well as plasmid sequencing for validation purposes and as described in the MATERIALS AND METHODS. Electrophoretic examination of the plasmid preparation (Figure 1 ) clearly showed that plasmid DNA, presumed to be pNZ44+100-ECas9i, had been introduced to BIM100-E-D1A-L5 and BIM100-E-D1A-L7 compared to the parent and original BIMs which showed no native or introduced plasmids. The MCS of pNZ44 in both the original plasmid and antisense constructs were also amplified (Figure 2) and showed a clear difference in size, the difference corresponding to the size of the S. thermophilus 100-E cas9 gene. The purified plasmid DNA was subsequently subjected to sequencing as described in the MATERIALS AND METHODS, and showed the addition of the S. thermophilus 100-E cas9 gene in the reverse orientation between the Pstl and Xbal restriction sites on the pNZ44 plasmid vector in both transformants. Finally, the CRISPR loci of BIM100-E-D1A-L5::pNZ44+100-ECas9i and BIM100-E-D1A- L7::pNZ44+100-ECas9i were amplified (Figure 4) and subjected to sequencing. The CRISPR spacer content of each transformant was 100% identical to each respective original BIM, confirming direct derivation.

The BIM100-E-D1A-L5::pNZ44+100-ECas9i and BIM100-E-D1A- L7::pNZ44+100-ECas9i transformants were subjected to phage sensitivity assays as described in the MATERIALS AND METHODS, the results of which are shown in Table 6. The results clearly show the suppression of acquired phage-resistance from BIM100- E-D1A-L7::pNZ44+100-ECas9i, despite the retention of the effective extra CRISPR spacer mentioned above. This suppression of resistance was not observed for BIM100- E-D1A-L5::pNZ44+100-ECas9i, highlighting the CRISPR1 independence of the BIM100- E-D1 A-L5 mechanism of phage-resistance.

Table 6. Relative efficiencies of plaquing (EOP) of phages of Streptococcus thermophilus strain 100-E, derived BIM and transformant.

The BIM100-E-D1A-L7::pNZ44+100-ECas9i transformant was then subjected to plasmid curing as described in the MATERIALS AND METHODS. Upon electrophoretic examination, the cured derivative showed an identical plasmid content of the parent S. thermophilus 100-E and original BIMs (i.e. no native or introduced plasmids; Figure 1 ). The cured derivative was then subjected to phage sensitivity assays as described (Table 6), the results of which show the restoration of acquired resistance to φΙ ΟΟ-Ε-DIA-L of the original BIM100-E-D1A-L7.

1.3 - Reduction in CRISPR-mediated BIM incidence in 100-E BIM generation

In order to reduce the incidence of CRISPR spacer alterations during BIM generation, S. thermophilus 100-E (parent) was transformed with freshly isolated pNZ44+100-ECas7i or pNZ44+100-ECas9i DNA, and Cm-resistant transformants were screened as described in the MATERIALS AND METHODS. BIM generation was performed, presumed BIMs purified and CRISPRs 1 and 3 amplified as described in the MATERIALS AND METHODS. CRISPR2 was omitted from this experiment due to its apparent redundancy in S. thermophilus, and only the leader end of each locus was sequenced due to the significant bias in iterative spacer incorporation in the remaining two CRISPR systems (Horvath et al., 2008). Sequencing was performed from the leader end of the CRISPR 1 and 3 loci of a total of 90 presumed BIMs (30 derived using 100-E parent, 30 using 100-E:: pNZ44+100-ECas7i and 30 using 100-E:: pNZ44+100-ECas9i - 10 BIMs each being derived from three separate experiments), using the forward primers shown in Table 2 (Seq IDs 1 , 7, and 9). At least 800 bp (CRISPR1 ) or 500 bp (CRISPR3) was sequenced for each potential BIM. Two CRISPR3 sequences were omitted due to poor data. The results of the cas-silenced BIM generation experiment are summarised in Table 7.

Table 7. Summary of CRISPR alterations in in presumed BIMs generated using S. thermophilus 100-E parent and cas7 and cas9 silenced strains.

It is clear from the above results that there is a significant reduction in the incidence of CRISPR alterations in derived BIMs of 100-E using both the cas7 and cas9 silenced transformants. It should be noted that both silencing constructs are designed to act on cas genes associated with CRISPR1 - for this reason some transference of activity to CRISPR3 (as shown by the slight increase in CRISPR3 alterations) can be explained.

1.4 - Generation of a 'double hurdle' BIM of S. thermophilus 100-E

In order to obtain BIMs of 100-E that are phage resistant due to both the mechanism of action of the CRISPR system and by a presumed mutation in a component of the cell envelop that acts as a phage receptor, so called 'double hurdle' (DH) BIMs of 100-E were generated (BIM100-E-DH51 and BIM100-E-DH61 ; Table 1 ). BIM100-E-D1A-L7 was considered to be a 'pure' CRISPR-mediated BIM, owing to the effectiveness of the CRISPR silencing method in restoring sensitivity to cp100-E-D1A-L (Table 6). For this reason, BIM100-E-D1A-L7 was applied in the generation of the DH BIMs.

It is clear from Table 6 that the CRISPR mechanism provides BIM100-E-D1A-L7 with resistance against cp100-E-D1A-L only, while cp100-E-D2A-L, cp100-E-D3A-L and cp100-E-D4A-L retain the ability to infect the BIM. This sensitivity was exploited during DH BIM generation. BIM100-E-D1A-L7 was firstly transformed with the Cas7 silencing plasmid, pNZ44+100-ECas7i (Table 3; Figure 7), to increase the frequency of secondary non-CRISPR BIM selection. The BIM was then subjected to a phage challenge using cp100-E-D2A-L and a number of potential BIMs were purified as described in the MATERIALS AND METHODS. Selected potential BIMs were screened on the basis of phage sensitivity and CRISPR sequences as well as sedimentation profile and chain length which are indicators of potential cell wall mutation-harbouring BIMs. Double Hurdle BIMs were subjected to plasmid curing (Figure 7) as described above prior to phage sensitivity assays.

Table 8: Summary of CRISPR repeat/spacer loci content in S. thermophilus strain

Table 9. Relative efficiencies of plaquing (EOP) of phages of Streptococcus thermophilus strain 100-E and derived BIMs.

Phage

Parent BIM

<p100-E-D1A-L <p100-E-D2A-L <p100-E-D3A-L <p100-E-D4A-L

100-E (parent) 1 1 1 1

BIM100-E-D1A-L7 8.67±0.15 x 10 ^~8 0.64 ± 0.13 0.85 ± 0.17 0.51 ± 0.04

BIM100-E-DH51 < 10 ^"9 1.93 ± 1 .32 x 10 ^~4 2.19 ± 1 .08 x 10 ^~5 < 10 ^"6

BIM100-E-DH61 < 10 ^"9 3.21 ± 1.02 x 10 ^"4 2.71 ± 1 .73 x 10 ^~5 < 10 ^"6 Note 1 : < denotes the limit of detection i.e. no plaques were detected when the derived BIM was challenged with a phage lysate containing 1 x 10 ⁹ pfu/ml phages.

Note 2: In derived BIM nomenclature, D1A-L denotes the phage against which the BIM was generated.

Table 10. Relative cells per chain (CPC) of parent and BIMs of S. thermophilus strain 100-E.

BIM100-E-DH51 and BIM100-E-DH61 can be defined 'double hurdle' BIMs for two reasons. Firstly, all three CRISPR repeat/spacer loci sequenced were found to be 100% identical to BIM100-E-D1A-L7 (Table 8), combined with the apparent absence of any additional CRISPR-Cas loci on the genome. This indicates that a mechanism other than the CRISPR system provided the additional resistance to phages cp100-E-D2A-L, φ100- E-D3A-L and cp100-E-D4A-L which was not observed in BIM100-E-D1A-L7. In addition, the sedimentation profile (Figure 8; Table 1 1 ) and relative increase of cell chain length (Figure 6; Table 10) indicate that these BIMs have undergone a mutation in cell envelope component that acts as a phage receptor. The combination of the CRISPR and non-CRISPR 'hurdles' to phage infection in one strain in this case leads to the definition of double hurdle BIMs. Table 1 1. Optical densities of S. thermophilus 100-E and derived BIMs overnight broth cultures.

1.5 - Heap Lawrence assy of a 'double hurdle' BIM of S. thermophilus 100-E

To test the phage robustness of a provided CRISPR BI M, a non CRISPR BI M and a double hurdle (DH) BI M's, the BI M's were subjected to a Heap Lawrence assay against phage q>100-E-D1 A-L as described in the MATERIALS AND METHODS, the results of which are shown in table 12 below.

Table 12. summary Heap Lawrence data against phage <p100-E-D1A-L

It is clear from the above results that an increased phage robustness is obtained for a non CRISPR BI M BIM 100-E-D1 A-L5 in view of CRISPR BI M BI M100-E-D1A-L7. Further, the above results show that a double hurdle BI M BIM 100-E-DH51 and BI M 100- E-DH61 provide a further increased phage robustness. In other words, the combination of different phage resistant mechanisms works synergistically by increasing the phage robustness to phage cp100-E-D1A-L. Example 2

CRISPR silencing of Parent and Bacteriophage Insensitive Mutants (BIMs) of S. thermophilus 100-F 2. Silencing of cas9 in S. thermophilus 100-F BIMs

2.1. Generation and analysis of 100-F CRISPR BIMs

Bacteriophages against S. thermophilus 100-F (3 in total; Table 1 ) and BIMs against these phages were isolated as described in the MATERIALS AND METHODS. One BIM, S. thermophilus BIM100-F-NGBD1A-L2 was selected for further analysis by plaque assay and CRISPR sequencing, the results of which are shown in Tables 13 and 14, respectively.

Table 13. Relative efficiencies of plaquing (EOP) of phages of Streptococcus thermophilus strain 100-F and derived BIMs.

Table 14. Summary of CRISPR repeat/spacer loci content in S. thermophilus strain 100-F and derived BIMs.

#

Size

Parent BIM CRISPR Direct repeat spacer Terminal repeat

(bp) s

5'- 5'-

100-F (parent) 2147 GTTTTTGTACTCT 32 GTTTTTGTACTC

CAAGA I I I AAGT TCAAGA I I I AAG

1

AACTGTACAAC-3' TAACTGTACAGT

BIM100-F-

2212 33 -3' NGBD1A-L2 SEQ ID. No 30 BIM100-F- SEQ ID. No 31 NGBD1A- L2::pNZ44

BIM100-F- NGBD1A- L2::pNZ44+100- FCas9i

BIM100-F- NGBD1A-L2

(cured)

100-F (parent)

BIM100-F- NGBD1A-L2 5'- 5'-

GATATAAACCTA GATATAAACCTA

BIM100-F- ATTACCTCGAGA ATTACCTCGAG NGBD1A- 2 258 GGGGACGGAAA AGGGGAC I I I I L2::pNZ44 3

C -3' TT-3

BIM100-F- NGBD1A- SEQ ID. No 32 SEQ ID. No 33 L2::pNZ44+100- FCas9i

BIM100-F- NGBD1A-L2

(cured)

100-F (parent)

BIM100-F- NGBD1A-L2 5'-

Ij 1 1 1 1 AIJAIJU I

BIM100-F- TGTTG 1 1 1 CGAAT

NGBD1A- 12

3 G GTTCCAAAAC - L2::pNZ44 827 As direct repeat

3'

BIM100-F- NGBD1A- SEQ ID. No 34

L2::pNZ44+100- FCas9i

BIM100-F- NGBD1A-L2

(cured)

100-F (parent) 5'- 5'-

GTTTTTCCCGCA GTTTTTCCCGC

4 762 CACGCGGGGGT 12 ACACGCGGGG

BIM100-F- GATCC - '3 GTGATTC - 3' NGBD1A-L2 SEQ ID. No 35 SEQ ID. No 36 BIM100-F- NGBD1A- L2::pNZ44

BIM100-F- NGBD1A- L2::pNZ44+100- FCas9i

BIM100-F- NGBD1A-L2

(cured)

The above results indicate that the observed insensitivity of BIM100-F-NGBD1A-L2 to at least two of the three 100-F phages was conferred by the CRISPR mechanism, due to the addition of one spacer at the leader end of the CRISPR1 repeat/spacer locus. This 30 bp spacer showed 100% homology to a region in a hypothetical protein in the replication modules of phages cpSTV84-D1A-L and cpNGB-D1 A-L (which was used in the challenge). The genome of cpSTV88-D1A-L was not available for analysis in this manner.

2.2. Transformation, curing & phage sensitivities of BIM100-F-NGBD1 A-L2

S. thermophilus BIM100-F-NGBD1A-L2 was transformed using freshly isolated pNZ44 and pNZ44+100-FCas9i, and Cm-resistant transformants were screened and then checked using the confirmatory PCR as described in the MATERIALS AND METHODS. Two presumptive transformants, BIM100-F-NGBD1A-L2::pNZ44 (plasmid control) and BIM100-F-NGBD1A-L2::pNZ44+100-FCas9i (proof of principle) were selected for validation and analysis.

BIM100-F-NGBD1A-L2::pNZ44 and BIM100-F-NGBD1A-L2::pNZ44+100-FCas9i were subjected to plasmid preparation followed by sequencing, pNZ44 MCS PCR (Figure 1 1 ), CRISPR PCR followed by sequencing (Table 14) and plasmid sequencing for validation purposes and as described in the MATERIALS AND METHODS. PCR amplification of the pNZ44 MCS regions of all strains (Figure 12) clearly showed that an approximately 500 bp product, presumed to be the MCS region of pNZ44 had been introduced to BIM100-F-NGBD1A-L2 compared to the parent original BIM which showed no native or other introduced plasmid products. The MCS of pNZ44 in both the original plasmid and antisense constructs were also amplified (Figure 1 1 ) and showed a clear difference in size, the difference corresponding to the size of the S. thermophilus 100-F cas9 gene. The purified plasmid DNA was subsequently subjected to sequencing as described in the MATERIALS AND METHODS, and showed the addition of the S. thermophilus 100-F cas9 gene in the reverse orientation between the Pstl and Xbal restriction sites on the pNZ44 plasmid vector in both transformants. Finally, the CRISPR loci of BIM100-F-NGBD1A-L2::pNZ44 and BIM100-F-NGBD1A-L2::pNZ44+100-FCas9i were amplified and subjected to sequencing. The CRISPR spacer content of each transformant was 100% identical to BIM100-F-NGBD1A-L2, confirming direct derivation. This was also confirmed by GTG RAPD PCR profiling (Figure 10).

The BIM100-F-NGBD1A-L2::pNZ44 and BIM100-F-NGBD1A-L2::pNZ44+100- FCas9i transformants were subjected to phage sensitivity assays as described in the MATERIALS AND METHODS, the results of which are shown in Table 15. The results clearly show the suppression of acquired phage-resistance of BIM100-F-NGBD1A- L2::pNZ44+100-FCas9i, despite the retention of the CRISPR spacer mentioned above. This suppression of resistance was not observed for phage cpSTV88-D1A-L Due to the the unavailability of the complete genome sequence of this phage, it was not possible to confirm the presence of identical protospacer sequence on the genome. Considering the absence of any additional spacers in the CRISPR2, CRISPR3 or CRISPR4 loci, it is probable that the resistance of BIM100-F-NGBD1A-L2 to cpSTV88-D1A-L is conferred by a mechanism other than CRISPR.

Table 15. Transformation, curing & phage sensitivities of CRISPR and silenced CRISPR BIMs of 100-F

Phage

< < <

Pare nt/B I M/Tra nsf o rma nt a 9

■4 9

m

> > o

H H z

CO CO

9- 9- 9-

100-F (parent) 1 1 1

4.04 x 10 ^"6 ± 3.97 x 10 ^"6 ±

BIM100-F-NGBD1A-L2 < 10 ^~7

6.41 x 10 ^~7 2.13 x 10 ^"6

6.04 x 10 ^"6 ± 5.95 x 10 ^"6 ±

BIM100-F-NGBD1A-L2::pNZ44 < 10 ^~7

5.49 x 10 ^"7 1.42 x 10 ^"6

1.02 x 10 ^"5 ±

BIM100-F-NGBD1A-L2::pNZ44+100-FCas9i 0.08 ± 0.07 0.03 ± 0.007

9.34 x 10 ^~7

BIM100-F-NGBD1A-L2::pNZ44+100-FCas9i 1.65 x 10 ^"5 ± 5.5 9.52 x 10 ^"7 ± 4.07 x 10 ^"6 ±

(Cured) x 10 ^"6 1.35 x 10 ^"6 2.92 x 10 ^"6 The BIM100-F-NGBD1A-L2::pNZ44+100-FCas9i transformant was then subjected to plasmid curing as described in the MATERIALS AND METHODS. Upon electrophoretic examination of pNZ44 plasmid MCS PCR products, the cured derivative showed an identical profile of the parent S. thermophilus 100-F and original BIM100-F- NGBD1A-L2 (i.e., no product; Figure 1 1 ). The cured derivative also showed an identical RAPD PCR pattern to its parent BIM (Figure 10). The cured derivative was then subjected to phage sensitivity assays as described (Table 15), the results of which show the restoration of acquired resistance to cpSTV84-D1A-L and φΝΘΒ-DIA-L of the original BIM100-F-NGBD1A-L2.

2.3 - Simultaneous Silencing of multiple cas genes to achieve reduced CRISPR- mediated BIM incidence during 100-F BIM generation S. thermophilus 100-F was previously observed to produce a high frequency of

CRISPR-mediated BIMs during BIM generation. In order to obtain non-CRISPR- mediated BIMs, the incidence of CRISPR-mediated BIMs in this strain had to be reduced. To this end, plasmids pNZ44+100-FCas7i, pNZ44+100-F2Casi and pNZ44+100-F2CasiSW strain were individually introduced into strain 100-F by electrotransformation and transformants were screened in each case for Cm-resistance, and then verified as described in the MATERIALS AND METHODS. All strains showed an identical GTG RAPD PCR profile (Figure 12), confirming direct derivation. The results of a pNZ44 MCS confirmatory PCR (Figure 13) clearly show the introduction of plasmids into respective transformants, with PCR product size correlating with insert size in all cases. BIM generation was performed using these transformants, presumed BIMs purified and CRISPRs 1 , 3 and 4 amplified as described in the MATERIALS AND METHODS. As was described for strain 100-E, CRISPR 2 was omitted from this analysis due to its apparent redundancy in S. thermophilus, and only the leader end of each CRISPR locus was sequenced due to the significant bias in iterative spacer incorporation in the remaining two (and, it can be assumed, CRISPR 4) systems (Horvath et al., 2008). Sequencing was performed from the leader end of the CRISPR 1 , 3 and 4 loci of a total of 46 presumed BIMs (10 derived from the 100-F parent, 10 from the control strain 100-F:: pNZ44, 6 from the 100-F:: pNZ44+100-FCas7i strain, 10 from the 100-F::pNZ44+100-F2Casi strain, and 10 from the 100-F::pNZ44+100-F2CasiSW strain), using the forward primers shown in Table 2 (Seq IDs 1 , 9 and 18). At least 400 bp of each locus was sequenced for each potential BIM, which, combined with GTG RAPD profiling (above), served to confirm direct derivation of each transformant (and BIM) from the parent 100-F strain. The results of the cas-silenced BIM generation experiment are summarised in Table 16.

Table 16. Summary of CRISPR alterations in presumed BIMs generated using S. thermophilus 100-F parent and cas7, 2Cas and 2CasSW silenced strains.

In general, it is noteworthy that while BIMs could readily be obtained from 100-F, 100-F plus pNZ44 (plasmid control), 100-F containing plasmid pNZ44+100-F2Casi, and 100-F harbouring plasmid pNZ44+100-F2CasiSW (10 BIMs obtained in approximately 2 rounds of BIM generation), 100F harbouring the silencing plasmid pNZ44+100-FCas7i produced BIMs at a lower frequency which necessitated at least 4 rounds of BIM generation, yielding a total of just 6 BIMs. A possible explanation for this may be the dominance of CRISPR1 activity in 100-F. A dedicated (i.e. single target) antisense construct such as pNZ44+100-FCas7i may lead to a substantial reduction in CRISPR1 activity and hence to a markedly reduced frequency of generated BIMs.

Despite this reduced frequency of BIMs, it is clear from the above results that there is a significant reduction in the incidence (if not a total absence) of CRISPR1 alterations in the examined BIMs obtained when challenging 100-F harbouring the cas7 silencing plasmid. Similar to the results seen for 100-E above (Table 7), transference of activity to CRISPR3 was noted when the CRISPR1 system was effectively disabled by silencing Cas7. In addition, two BIMs were obtained which showed no alterations in CRISPR1 , 3 or 4 loci. BIM generation using the 100-F::pNZ44+100-F2Casi did not appear to have a significant effect on the incidence of CRISPR-mediated BIMs. A possible explanation for this is the location of cas7 (antisense) on the silencing vector - having been cloned behind csn2, its level of effective anti-sense transcripts may have been negatively affected due to, for example, mRNA instability. Considering that spacer addition in CRISPR1 is the most likely adaptation in BIMs generated from phage-challenged 100-F (as evidenced by the proportion of CRISPR1 -mediated BIMs derived from 100-F), it may be more effective to silence cas7 by locating it its anti-sense orientation next to the p44 promoter on the silencing construct. This was attempted by switching the locations of cas7 and csn2 in the silencing vector pNZ44+100-F2CasiSW. By using this construct, five BIMs which showed no spacer additions in CRISPR1 , 3 or 4 were generated (of ten analysed).

Applicant's or agent's file reference number 30911-WO-PCT International application No.

INDICATIONS RELATING TO A DEPOSITED MICROORGANISM

(PCT Rule Ubis)

The indications made below relate to the microorganism referred to in the description

first mentioned on page 11 line 10.

B. IDENTIFICATION OF DEPOSIT Further deposits are identified on an additional sheet

Name of depositary institution

CENTRAAL BUREAU VOOR SCHIMMELCULTURES

Address of depositary institution (including postal code and country)

Uppsalalaan 8

P.O. Box 85167

NL-3508 AD Utrecht

The Netherlands

Date of deposit 15 July 2014 Accession Number CBS 138555

ADDITIONAL INDICATIONS (leave blank if not applicable) This information is continued on an additional sheet |

We inform you that the availability of the microorganism identified above, referred to Rule 13bis PCT, shall be effected only by issue of a sample to an expert nominated by the requester until the publication of the mention of grant of the national patent or, where applicable, for twenty years from the date of filing if the application has been refused, withdrawn or deemed to be withdrawn.

D. DESIGNATED STATES FOR WHICH INDICATIONS ARE MADE (if the indications are not for all designated States)

E. SEPARATE FURNISHING OF INDICATIONS (leave blank if not applicable)

The indications listed below will be submitted to the International Bureau later (specify the general nature of the indications e.g., "Accession Number of Deposit ")

For receiving Office use only For International Bureau use only

This sheet was received with the international This sheet was received by the International Bureau application on:

Authorized officer Authorized officer

Ku i per-Cristi na, Nathal ie

Form PCT/RO/134 (July 1992) Applicant's or agent's file reference number 30911-WO-PCT | International application No.

INDICATIONS RELATING TO A DEPOSITED MICROORGANISM

(PCT Rule Ubis)

The indications made below relate to the microorganism referred to in the description

first mentioned on page 12 line 3.

B. IDENTIFICATION OF DEPOSIT Further deposits are identified on an additional sheet

Name of depositary institution

CENTRAAL BUREAU VOOR SCHIMMELCULTURES

Address of depositary institution (including postal code and country)

Uppsalalaan 8

P.O. Box 85167

NL-3508 AD Utrecht

The Netherlands

Date of deposit 6 May 2015 Accession Number CBS 139996

ADDITIONAL INDICATIONS (leave blank if not applicable) This information is continued on an additional sheet |

We inform you that the availability of the microorganism identified above, referred to Rule 13bis PCT, shall be effected only by issue of a sample to an expert nominated by the requester until the publication of the mention of grant of the national patent or, where applicable, for twenty years from the date of filing if the application has been refused, withdrawn or deemed to be withdrawn.

D. DESIGNATED STATES FOR WHICH INDICATIONS ARE MADE (if the indications are not for all designated States)

E. SEPARATE FURNISHING OF INDICATIONS (leave blank if not applicable)

The indications listed below will be submitted to the International Bureau later (specify the general nature of the indications e.g., "Accession Number of Deposit ")

For receiving Office use only For International Bureau use only

This sheet was received with the international This sheet was received by the International Bureau application on:

Authorized officer Authorized officer

Ku i per-Cristi na, Nathal ie

Form PCT/RO/134 (July 1992)