Technical Field

The invention relates to the use of genetic polymorphism detection to diagnose presence or probabil¬ ity of disease. More particularly, it relates to the use of these techniques to assess the risk of, and to follow familial patterns of, the incidence of high renin hypertension, hypertension of other types, and variable, plasma renin activity. Other correlations are also dis¬ closed, and the pattern of genetic polymorphisms is also a tool for genetic identification.

Background Art

Hypertension is one of the most important pub- lie health problems in developed countries. It affects more than 60 million people in the United States alone. The disease, which is defined as an elevation of arte¬ rial blood pressure, results in secondary organ damage and a reduced..1ifespan. A significant form of hypertension is associ¬ ated with; art array-of symptomology indicating the involvement of the aldosterone/renin/angiotensin system. There is a series of metabolic interactions that are believed to characterize this form of hypertension. The measured hypertension itself, the increased blood pres¬ sure measured in the arterial system, has, as its imme¬ diate cause, vasoconstriction effected by angiotensin II. Angiotensin II is the product of metabolically inactive precursors which are activated by renin. Renin

is secreted by the kidneys, and also causes the release of aldosterone from the renal cortex. The levels of excretion of various ions and compounds are also affected. Whatever their precise interrelationship, indicators related to this mechanism are known to be associated with the condition designated high renin hypertension. These indicators include, in addition to high blood pressure itself, elevated levels of plasma renin activity (PRA) , elevated levels of urinary aldosterone (UA) , diminished levels of urinary sodium ion (U(Na)V), and increased levels of urinary potassium (U(K)V). Elevated PRA is always associated with the disease; the remaining parameters appear more secondary. There are, of course, other types of hyperten¬ sion which do- not follow this pattern. Also, some of these indicators, when correlated independently, are diagnostic of other conditions.

There is, at present, no predictive diagnosis for high renin hypertension or of hypertension in gen¬ eral. It is singularly important to have a specific predictive -tool. in.the case of this disease, as various preventive therapies are available, and the choice of appropriate preventive therapy depends not only on rec- ognition of the potential for hypertension per se, but also on correct ^" assessment of- the type of hypertension predicted. For example, for high renin hypertension, but not for other types, administration of Captopril® or Enalapril®, inhibitors of- angiotensinconverting enzyme, is sometimes helpful. In relatively benign cases, but also depending somewhat on the metabolic basis for the hypertensive symptoms, even regulation of diet, lifestyle, or other noninvasive practices are effective in minimizing risk.

A technique that inherently offers the advan¬ tages of early detection, if its results can be effec¬ tively correlated with the disorder to be assessed, is genetic analysis. Since the genomic characteristics of an individual are basically determined, it is assumed, at conception, genetic aberrations which are indicia of later metabolic disorders are an ideal early diagnosis tool. Genetic testing can be routinely conducted using present methodology, as early as the seventh week of fetal life. In adults and children, it can be accom¬ plished using a small blood sample. Over the last ten years, the availability of restriction enzymes and DNA probing techniques has made possible the use of "restriction fragment length polymorphisms" (RFLPs) in such diagnosis. Using the, by now, well established Southern blot hybridization technique (Southern, E., J Mol Biol (1975) 98.:503-517) , it has been possible suc¬ cessfully to diagnose sickle cell anemia (Kan, Y.W., et al, Proc Natl Acad Sci (USA) (1978) 7_5_:5631); B-thalassemia (Antonarakis, S.E., et al, Proc Natl Acad Sci (USA) (1983) 21:137); type II diabetes (Rotwein, P., et al, Science (.1981) ^' 213:1117) ; familial growth hormone deficiency (Phillips, J.A., III, Banburv Report 14, Cold Spring Harbor- Laboratory (1983) pp 305-315); phenylketonuria (Woo, S.L.C., et al, Nature (1983)

306:151) ; Huntington's disease (Gusella, J.F., et al, Nature (1983) 306:234) : and hemophilia B (Gianelli, et al, Lancet (1984) i.:239, Grunenbaum, et al, J Clin Invest (1984) 21:1491). The use of genetic detection systems is par¬ ticularly appropriate in predicting the incidence of hypertension for an individual since it appears that genetic, rather than environmental, factors are in large part responsible for the onset of the disease. This has

been shown in both animal and human studies, including of blood pressures which exist as patterns or "aggrega¬ tions" within families. Adopted children do not demon¬ strate familial aggregation of blood pressures. There- fore, an assessment of the genome of the subject pro¬ vides fertile ground for assessment of risk.

All of the successful predictive tests listed above are grounded in identification of a particular polymorphism or polymorphisms which correlates with the disease or disorder in question. It has been calculated that the number of polymorphisms expected in the human genome should be of the order of 10 , based on an assumed probability of 0.005 for a given nucleotide to be polymorphic; this number being inferred from studies of the human growth hormone, αl-antitrypsin and B-like globin gene cluster loci (Jeffreys, A. J., Cell (1979) 1£:1-10; Oster, H. , et al, Am J Hum Gen (1984) 36(SUPP1) 150S). The challenge is to determine which of these over ten million polymorphisms is associated with a par- ticular disorder, and to devise a procedure to detect it.

The-present ^* invention, in part, uses the L/M/U insertion upstream of the insulin gene as a test site for the occurrence of polymorphisms associated with high renin hypertension.'

There-- is~ ncertainty in the literature as to whether a 1600-2200 bp ("U") insertion polymorphism, which has a 0.32 frequency in the population, does (Owenbach, D.G., et al, Lancet (1982) 1291-1293; MandrupPoulsen, T., Lancet (1984) 250-252) or does not (Jowett, N.I., et al, Lancet (1984) 348) correlate to atherosclerosis. The "U" insertion correlates well with an increased risk, and a shorter 600-1600 bp "M" inser¬ tion (freq. = 0.04) correlates moderately with a

decreased risk of atherosclerosis as set forth in U.S. Serial No. 900,593, filed 26 August 1986, assigned to the same assignee and incorporated herein by reference. Also, Canadian patent 1,186,991 discloses correlations with atherosclerosis for the presence of the U insertion and with non-insulin-dependent diabetes for the absence of the two insertions.

The gene encoding renin also shows useful polymorphisms. Appropriate DNAs encoding human renin ^' are also available. Human renin cDNA and genomic DNA have been described by Hardmann, J.A., et al, DNA (1984) 2:457; Imai, T. , et al, Proc Natl Acad Sci (USA) (1983) ___: 7405; Hobart, P.M., et al, Proc Natl Acad Sci (USA) (1984) 8_1: 5026; Miyazaki, H. , et al, Proc Natl Acad Sci (USA) (1984) 8_1: 5999. Means for detection of the L/M/U insertion in the insulin region are known in the art.

Additional polymorphisms predictive of hyper¬ tension are in- the genes encoding atrial natriuretic peptide (ANP), the apoCIII and apoAI genes, and the estrogen receptor gene.

The present invention provides a number of polymorphic sites,.in the above-mentioned gene regions which are useful both alone and in combination in pre¬ dicting the incidence of high renin hypertension and of other forms of hypertension of variable PRA and of other metabolic parameters by means of a simple blood test. The pattern of the polymorphisms in an individual also provides a tool for characterizing the subject and for establishing familial relationships.

Disclosure of the Invention

The invention provides identification of a polymorphism which is predictive of the subsequent development of high renin hypertension, and of

additional polymorphisms which are predictive for hyper-tension per se, variable PRA, and other metabolic indications, and which contribute to the genetic iden¬ tity of individuals and families. Thus, in one aspect, the invention is directed to a method of predicting the likelihood of development of high renin hypertension in an individual, which method comprises detecting the presence or absence of the U-allele insertion of the insulin gene. ("Polymorphism" refers, in the context of the present

-invention, to a location in the DNA of the genome which is different in various individuals. Usually, for a given location in the genome where such polymorphism occurs, a majority of alleles present in a population of individuals will be of one form, and a minority of the alleles present in the population will be of another. For convenience, and for want of a better general term, the form found in the minority of alleles will be referred to as "the polymorphism", using the form pre- sent in the majority as a reference. This is in analogy to the manner in which, for example, the term "isotope" is often used.vto.refer to the less common nuclides of an element. )

Some forms of hypertension per se, associated with high systolic or diastolic blood pressure or both, - are correlated with the presence of one or more of the Bglll polymorphism of the renin gene, the TaqI polymorphism of the renin gene, the HindiII polymorphism of the renin gene (which has a negative correlation with high diastolic blood pressure), the Bgll polymorphism of the ANP gene (which correlates with high systolic blood pressure), and the Bglll polymorphism of the ANP gene (which correlates with both) . Ah apoAI deletion and a PvuII polymorphism in the apoCIII gene also show

correlations. Estrogen receptor gene polymorphisms may show correlations as well.

Other correlations include correlation with variable PRA and urinary aldosterone levels for the Bglll and TaqI renin polymorphisms, with urinary aldosterone levels for the Bgll ANP polymorphism, and with urinary sodium and potassium levels for the Bgll ANP and HindiII renin polymorphisms, as well as U(Cr)V for several of the above. The polymorphisms of the estrogen receptor gene are expected to correlate with the development of osteoporosis.

In still another aspect, the invention relates to identification of individuals or familial patterns using polymorphisms. Thus, in another aspect, the invention relates to identification of individuals or familial patterns, and/or establishing the risk of hypertension in subjects using polymorphisms of the renin, ANP, apoAI/CIII gene complex, estrogen receptor gene or the Na,K-ATPase genes.

Brief Description of the Drawings

Figure 1 shows the DNA sequence of the renin 5' and renin 3' probes. Figure 2 shows typical results in assessing the presence or absence of the Bgll polymorphism of the renin gene.

Figure 3 shows the "clone 6" probe used for detection of the polymorphisms in the ANP gene.

Modes of Carrying Out the Invention

In the description below, distances of polymorphisms from reference points and lengths of dele¬ tions are given, when known, in bp or kb. Where the sequence.is known, such measures can be quite precise,

but when assessed by measuring fragment sizes on gels or by other experimental means, these measures contain a margin of uncertainty, as is well understood in the art. In general, for measures of 4 kb, the margin of uncer- tainty is + ~ 0.3 kb; for smaller lengths, the error is + ~ 10%.

A. Techniques for Detection of Polymorphisms

Application of the method of the invention employs standard techniques of DNA extraction, purifica¬ tion, restriction enzyme digestion, and size separation. Techniques for hybridization with probe and detecting successfully hybridized substrate arranged according to molecular" weight are also well known to those in the art. The general approach to finding and detecting the significant ^* polymorphisms is the following:

DNA is extracted from the somatic cells of the individual to be tested, for example from leukocytes, placental cells, cultured fibroblasts, or, in the case of fetal individuals, from cells of the amniotic fluid. In ordinary circumstances, the test material comprises leukocytes-from a.-blood sample -25-50 ml blood yields 200-400 μg DNA. The high molecular weight DNA fraction is separated, and-subjected to treatment with a particu- lar, selected restriction enzyme, such as, for example, EcoRI, Ba HI, MstI, XmnI, and the like. After digestion of the high molecular weight DNA, the digest is applied to a polyacrylamide or agarose gel and subjected to electrophoresis to obtain separation of the DNA frag- ments resulting from restriction enzyme digestion into positions on the gel determined by the size (length) of the fragment. The contents of the gel are then repli¬ cated by transferring to a nitrocellulose filter or other suitable matrix for use as a probe hybridization

support. The DNA fragments, either before or after transfer to the nitrocellulose filter replica, are treated with a denaturant such as sodium hydroxide/salt. The denatured, single-stranded DNA, replicated electrophoresis patterns are probed with labeled (usu¬ ally by ³² P) single-stranded DNA fragments. Other labels besides radioactivity, such as fluorescent molecules, may also be used, or the presence of doublestranded (as opposed to single-stranded) DNA can be detected. Depending on the probe selected, fragments will be detected which derive from a particular- region on the genome. For example, in the method of the inven¬ tion to predict high renin hypertension, a cDNA sequence from the renin or insulin gene sequences is used as a probe. Therefore, the only fragments which will appear on the hybridized filters are those which contain sequences complementary to this probe—i.e., only those which have not'been severed either in the genome itself or by the restriction enzyme cleavage from the cσmple- mentary renin or insulin fragment. Stated in another way, by using a particular probe, alterations in the genome whic -are.proximal to sequences corresponding to that probe are detected.

The specific procedures used in the general process described in the preceding paragraphs are under¬ stood in the art. < Procedures for DNA extraction from somatic cells may, for example, be found in Kan, Y.W., et al, Proc Natl Acad Sci (USA) (1978) 75:56315635; Taylor, J.M., et al, Nature (1984) 251:392.-393: and Kan, Y.W., et al, N Eng J Med (1977) 212: 080-1084. Further improvements which permit rapid extraction of the DNA are also disclosed by Law, D. G. , et al, Gene (1984) 2_8_:153-158. Techniques for size separation of the restriction enzyme treated DNA fragments are also

described in the foregoing references. Restriction enzyme digestions are generally standard in the art and are carried out under buffer, ionic strength, and tem¬ perature conditions which are specified by the anufac- turer of the particular restriction enzyme.

Transfer to nitrocellulose or other support and probing by prehybridization with nonspecific DNA followed by hybridization with labeled probe are also standard procedures disclosed, for example, in the fore- going references and by Southern, E., (supra). The sec¬ tion of the genome which is fingerprinted or otherwise subject to study using the results is, of course, depen¬ dent on the nature of the probe.

B. Nature of the Gene Regions and Probes Useful in the Invention

The fragment pattern obtained is diagnostic for a particular polymorphism if the probe selected is complementary to a DNA sequence sufficiently proximal to the polymorphism on the genome that it is not severed from the polymorphism by the restriction cleavage, and has a low probability of being segregated from the polymorphism by crossing over. Acceptable distance lim¬ its between the region of probe complementarity and the polymorphism are therefore arbitrary. Generally, probes which hybridize to DNA sequences within 10 kb upstream or downstream of the polymorphism give acceptable results. Occasionally, the pattern of restriction enzyme cleavage may place a distal probe hybridization site on a fragment irrelevant to the polymorphism. The closer the probe to the polymorphism, the greater the range of usable restriction enzymes.

Accordingly, as used herein, a probe which is a "substantial equivalent" to a specified probe is one

which hybridizes to the same fragment length in digests of DNA from individuals with a particular polymorphism when the same restriction enzyme is used. Thus, for the "Bglll" polymorphism of the renin gene, when Bglll digestion is used for diagnosis, the renin 5' and renin 3' probes (described below) are equivalent. For Hindlll digestion, they are not. (Of course, the designated probe can be modified in a trivial manner by being made longer or shorter or by selecting a slightly displaced sequence, and the modified probe will be a substantial equivalent.) cDNA probes can be labeled by nick translation using [ ³² P] dCTP and α [ ³² P] dGTP. Oligomer probes are labeled by kinasing. cDNA probes which are comple- mentary only to the exon regions of the gene and which span over intron regions are workable in the method of the invention. Other labeling methods can also be used so as to utilize a variety of detection techniques. For example, the cDNA can be extended with a polyA tail and hybridized for detection to a fluorescent material or an enzyme conjugated to polyT. Indeed, kits are commer¬ cially available- hich employ this general principle. Thus, in addition to radioactive labeling, fluorescence or enzyme activity, may be used as the ultimate means for sensing the hybridization. Also available are antibod¬ ies which bind selectively to double-stranded DNA, as opposed to singlestranded forms. Thus, the hybridization can also be detected by use of these anti¬ bodies and any desired labeling method. Renin is a critical protein in the angiotensin system, which controls vasoconstriction/dilation and is, of course, central to high renin hypertension. (Unex¬ pectedly, while correlation to hypertension is shown for some renin gene modifications, and with variable PRA,

there seems to be no correlation with high renin hyper¬ tension. Indeed, the polymorphisms disclosed herein are seen to be associated with low PRA, but still to show, in some cases, positive correlation with hypertension. In high renin hypertension, high PRA corre¬ lates to hypertension because of its role in the angiotensin system. The vasoactive agent is angiotensin-II, an octapeptide which acts directly to cons.trict vascular smooth muscle and to induce the release aldosterone from the adrenal cortex. The level of angiotensin-II is determined by the balance between its rate of inactivation by angiotensinases and its rate of formation in a two step process from angiotensinogen. The rate-limiting step in the conversion of angiotensinogen to angiotensin-II is the step catalyzed by renin—the conversion of angiotensinogen to angiotensin-I. The second step is mediated by a "con¬ verting enzyme''. -Renin is secreted in vivo by the juxtaglomerular cells of the kidney, and is synthesized as a prorenin precursor, which is then processed to give the active renin form. The involvement of renin in the regulation of ^*' -blood,pressure has been established by the utility of inhibitors of renin secretion or direct renin inhibitors. See, for example, Haber _r E., Hypertension and the Angiotensin System: Therapeutic Approaches (1984) Raven Press, 133-145; Gagnol, J.P., et al, Abstracts International Society of Hypertension (1984) 10:376.

Correlations have also been found, as further described below, between hypertension in general, and certain polymorphisms of the ANP, apoAI and apoCIII genes. In addition, polymorphisms of the estrogen receptor gene are expected to show correlations and Na,K-ATPase genes are expected to show correlations.

Steroid metabolism is known to influence blood pressure, Na,K-ATPase is a known Na ⁺ pump; abnormal Na ⁺ excretion is often associated with hypertension.

The insulin gene has no apparent relationship to high renin levels or to hypertension, but recent results, reported by Christlieb, A.R. , et al in Diabetes and Hypertension (1985) 2(Suppl 2):II-54 to 11-57, sug¬ gest that there is a causal relationship between the level of circulating insulin and diastolic blood pres- sure. These authors further suggest that because insu¬ lin increases the renal tubular reabsorption of sodium, it contributes to the hypertension. This result is also consistent with the finding of reduced levels of excre¬ tion for sodium in high renin hypertension. On the other hand, the L/M/U insertion upstream of the insulin gene, seems to have no effect on insulin production, and thus the correlation found herein would not have been predicted from any relationship of insulin levels to hypertension.

C. Methods to Obtain Correlations

Correlations of such polymorphisms with hyper¬ tension or other symptomology are analysed by applica¬ tion of suitable statistics. Two preferred methods are as follows:

In one method, the findings are interpreted in terms of the relative risk of persons having the polymorphism to show the disease, compared to those hav¬ ing an absence of the polymorphism. These "relative incidence" values are calculated according to Wolf, B., Ann Hum Genet (1955) _^ 9:251. As applied, for example, to the assays for propensity for hypertension, the rela¬ tive incidence is calculated as equal to:

PP x CN PN x CP

where PP is the number of hypertensive patients hav¬ ing the polymorphism;

PN is the number of hypertensive patients not having the polymorphism;

CP is the number of controls having the polymorphism;

CN is the number of controls not having the polymorphism.

The value calculated by this ratio, if signif¬ icantly greater than 1, ^" indicates that the persons hav- ing the polymorphism are at a greater risk of having hypertension; a value less than 1 shows protection against hypertension.

In another approach, the results of genetic analysis on persons known to have, for example, high renin hypertension and those known not to have it are correlated statistically according to the statistical method of Pearson,- described in standard statistics texts to obtain a "correlation coefficient", designated Rp, wherein Rp has a value in the range -1 (complete negative correlation), 0 (no correlation), and +1 (com¬ plete positive correlation). In practice, absolute val¬ ues of Rp above approximately 0.2 are considered suffi¬ cient to demonstrate good correlation between the param¬ eters measured. A "p" value is also calculated to show the degree of confidence in the Rp correlation coeffi¬ cient, according to the standard student "t" test; val¬ ues of "p" less than 0.05 indicate a 95% confidence level.

High renin hypertension is characterized by a pattern of parameters, including high renin concentra¬ tion in the plasma and the high blood pressure per se. Therefore, in exploring correlations with regard to high renin hypertension, not only blood pressure, but other parameters associated with high renin hypertension, were measured and correlated statistically as above described with the polymorphism in question. Of particular impor¬ tance, of course, is correlation with blood pressure per se and with high levels of plasma renin activity (PRA). Also measured were correlations with the other known characteristics of the disease—i.e., urinary sodium (U(Na)V) decrease, urinary potassium (U(K)V) increase and urinary aldosterone (UA) increase. Several indepen- dent correlations with these variables were found with regard to renin, ANP, ApoAI and ApoCIII polymorphisms. The L/M/U insulin polymorphism showed a complete pattern of correlations which demonstrates, dramatically, its nexus with high renin hypertension.

D. Kit Components ^"

The reagents suitable for applying the method of the invention to detect the appropriate polymorphisms may be packaged into, convenient kits providing the nec- essary materials, packaged into suitable containers, and, optionally,- suitable containers or supports useful in performing the assay. The essential components of the assay include the restriction enzyme associated with the polymorphism and a suitable probe. Additionally, packages containing concentrated forms of reagents used for hybridization, prehybridization, DNA extraction, etc. may be included, if desired. In particular, how¬ ever, labeled probe, or reagents suitable to form

conveniently labeled probe, are useful in facilitating the conduct of the method of the invention.

For example, for renin two probes have been used to detect polymorphisms in the region of the renin gene, the renin 5' and the renin 3' probe described below. Of course, other equivalent probes can be used, including simple modifications of the exemplified probes, which are constructed by shortening or lengthen¬ ing the probe sequences or by making minor alterations in the base sequence which do not significantly affect the hybridizing power of the probe. In addition, probes which represent DNA sequences which are slight displace¬ ments of the exemplified probes in the genome are also usable. The renin 5' probe represents the codons for the first 198 a ino acids of the mature renin sequence, and includes the codons for the amino acids of the presequence. ^" The renin 3' probe contains the remaining mature protein codons. It will be recalled that the nick translation method of labeling fragments the probe, and thus the presence of intron regions in the target gene does not undermine the effectiveness of the probes if this labeling method is used.

Ten polymorphisms of the renin region have been identified. These are designated Bgll, Bglll, Rsal, HindiII, TaqI, EcoRI, HphI, HinFI, EcoT14I and Mbol herein.

Thus, a kit for detection of renin polymor¬ phisms might include, in individual packages, the renin 5' and 3' probes, and the needed amounts of the above enzymes. The kit would also include instructions for DNA extraction, restriction cleavage, electrophoresis, and interpretation.

Examples

The following examples are specific with respect to the probes exemplified and with respect to the precise conditions of DNA extraction, probe hybridization, etc. It is understood that these factors are illustrative but not limiting. The essential fea¬ tures of the invention as it relates to detection of .a particular polymorphism are (1) the selection of enzyme, and (2) the selection of probe. For a particular restriction enzyme, a number of substantially equivalent probes which are not segregated from the identified fragment by this restriction cleavage are usable. Alternatively, other restriction enzymes may be used in conjunction with a probe which hybridizes in particu¬ larly close proximity to the polymorphism.

In the above sense, the particular restriction enzyme and cDNA probe chosen are arbitrary. However, it should be noted, as is understood in the art, that the efficacy of the probe is enhanced as it moves closer to the site of .the polymorphism. Otherwise, additional cleavage points- may be encountered between the polymorpHism and the probe, and also the probing site may be separated from the site of the polymorphism dur- ing replication by crossing-over events.

Procedures for Analysis

Leukocytes were obtained from freshly drawn blood collected from each of the human subjects, and high molecular weight genomic DNA was isolated by the procedure of Law, D. J., et al, Gene (1984) 28:153-158.

High molecular weight DNA was divided into portions and each was digested to completion with one of the various restriction enzymes under conditions

recommended by the suppliers (New England Biolabs and Bethesda Research Laboratories). The digests were electrophoresed in horizontal agarose gels in 30 mM NaH2P0 , 36 mM Tris, 1 mM EDTA, pH 7.7. After electrophoresis, DNA fragments were denatured .in situ in 0.5 M NaOH/1.5 M NaCl for 2 x 10 min, neutralized in 1 M ammonium acetate pH 7.2 for 2 x 10 min, and transferred overnight onto nitrocellulose paper (Schleicher and Schuell). The filters were rinsed in 2 x SSC (Ix SSC is 0.15 M NaCl, 0.015 M sodium citrate pH, 7.4) and baked for 2 hr at 80°C .in vacuo and then were prehybridized for 5 hr in plastic bags using 0.3 ml/cirr' of a solution containing 5x SSPE (Ix SSPE is 10 mM Na phosphate pH 7.4, 0.18 M NaCl and 1 mM EDTA) containing 5x Denhardt's solution (1 x Denhardt's contains 0.2 mg/ml each of

Ficoll, polyviπylpyrrolidone and bovine serum albumin), ^• 40% (vol/vol) formamide, and 250 μg/ml sheared and dena¬ tured salmon sperm -DNA, and hybridized overnight in the same bag in 0.1 ml/cm ² of 5x SSPE, lx Denhardt's solu- tion, 40% (vol/vol) formamide, 10% dextran sulfate, and 100 μg/ml sheared and denatured salmon sperm DNA, mixed with 100 ng per-bag (containing 1 or 2 filters) of the appropriate __ P-labeled probe, as discussed below.

Prehybridization and hybridization were performed at 42°C.

Filters were then washed twice at room temper¬ ature in 2x SSC and twice at 65°C in 2x SSC, lx

Denhardt's solution. DNA sequences hybridized to the

^* - _"? ^** ?labeled probes were visualized by autoradiography using XAR-5 films (Kodak) and Cronex intensifying screens (Dupont) at -70°C for 18 hr to 2 days.

The methods used to prepare probes are described; however, the complete DNA sequences of the probes used herein are given or are known so that

synthetic methods may also be used. Methods to obtain entire gene sequences, once the desired sequence is known, are no available. See, e.g., Edge, M.D. et al, Nature (1981) 292: 756; Nambiar, K.P., et al, Science (1984) 2_2_3: 1299; or Jay, E., et al, J Biol Chem (1984) 259: 6311.

For use in hybridization in these examples, the probes were labeled to a specific activity of 2.0-5.0 x 10 ⁸ cpm/μg by nick-translation, using the BRL nicktranslation kit (Bethesda Research Laboratories) under recommended conditions with α [ ³² P] dCTP (800 Ci/mM; Amersham Corporation) in the presence of unla- beled dATP and dTTP. The probes were denatured just before the hybridization step by incubation for 5 min in a boiling water bath, followed by rapid cooling in ice water.

Renin

Two specific probes are used for the renin region in the illustrations below: renin 5' probe and renin 3' probe, as shown in Figure 1.

To prepare the renin probes, a cDNA library was. prepared from oligo-dT-primed polyA ⁺ RNA, purified from the kidney of- a human accident victim. The cDNA library was constructed in the bacteriophage vector λgtlO and probed with appropriate fragments of Charon-4 human renin genomic clones which had been obtained from a human genomic library by probing with mouse submaxillary gland renin cDNA fragments. The λgtlO library was prepared by standard procedures which include obtaining double-stranded cDNA from the mRNA, blunt-ending the cDNA with DNA polymerasel (Klenow) , adding commercially available EcoRI linkers, cleaving with EcoRI, and ligating the fragments into

EcoRI-digested λgtlO vectors. Two positively responding cDNA clones together, when sequenced using the dideoxy method of Sanger (supra) , were shown to contain the entire renin-encoding sequence and 3' untranslated region except for a missing 7 base pairs at the 5' end of the signal sequence. The two clones include 1211 bp of coding region, followed by the entire 3' 198 bp untranslated region, followed, in turn, by a polyA tail. The assignment of the sequence to the appropriate preprorenin codons and reading frame was made by compar¬ ison to the published Imai et al (supra) sequence. Two cloned fragments result because of the EcoRI site shown in position 692 in Figure 1.

The cDNA sequences encoding the preprorenin. protein were transferred into pUC9 as the two EcoRI inserts to obtain the cloned vectors pHRl containing renin 5' probe and pHR2 containing renin '3 probe. pHRl and pHR2 were prepared by excising the cDNA from λgtlO using EcoRI and ligating each of the resulting fragments into EcoRI-cleaved pUC9. For use as probes, the inserts were removed with EcoRI and labeled.

Using the renin 5' and renin 3' probes, 68 individuals were tested for polymorphism by the above procedure. The following polymorphisms were found, as shown in Table 1.

Table 1

More Less

Poly Common Common morphism Probe Enzyme Eraqmeπt Fragment Frequency

Bgll renin 5' Bgll 5 kb 9 kb 0.39

Bglll renin 5' Bglll 20 kb 24 kb 0.28-0.30 renin 3'

Rsal renin 5' Rsal 0.9 kb 0.35 (US) renin 3' 0.40 (Germany)

Hindi11 renin 3' Hindlll 9.0kb 6.2kb 0.34 TaqI renin 5' TaqI 9.8kb ll.Okb 0.22-0.29 renin 3'

EcoRI mixture EcoRI 3.9kb 0.025+0.01 HphI renin 5' + HphI 0.25 0.74 0.10+0.03 renin 3'

HinFI renin 5" + HinFI 1.3 1.4 0.20+0.05 renin 3'

EcoT1 I renin 5'+ EcoT1 I 0.60 0.25 0.45+0.05 renin 3'

M ol renin 5'+ Mbαl 1.5 1.3 0.20+0.05 renin 3'

In the case of the HphI polymorphism detection is against a background of bands at 0.40, 0.45, 0.66, 0.90, 1.10 and 1.25 kb; for HmFI, of 0.35, 0.40, 0.44, 0.54, 0.64 and 0.85 kb; for EcoT14I, of 0.78, 1.10, 1.40, 2.20 and 3.00 kb; and for Mbol of 0.6, 0.8 and 2.0 kb.

Frequency in this case as in all others is determined as follows: Each individual has two copies' of each chromosome; therefore, at any locus or position on the chromosome each person has two "alleles". If these two alleles are identical,- the individual is "homozygous" at this locus; if the two alleles are dif¬ ferent, he is heterozygous at this locus. Genetic vari¬ ation between populations can be quantified using the concept, of allele.. frequency, the proportion of .all.

alleles in a population at the locus that are a particu¬ lar allele. The frequency of any particular allele in a sample is therefore equal to twice the number of homozygotes for that allele plus the number of heterozygotes for that allele divided by two times the number of individuals in the sample.

For example,- in the case of the Bgll (5 kb/9 kb) polymorphism there were:

— 8 homozygotes of the 5 kb type

—13 homozygotes of the 9 kb type —27 heterozygotes

The polymorphic allele frequency (9 kb type) is thus:

p = — ^{26 27} or approximately 0.39

56 + 26 + 54

Typical results are shown for the Bgll polymorphism in Figure 2. Each lane of the autoradiogram shows the results for one individual probed with renin 5' probe. Lanes 4, 7, 11 and 13 show individuals homozygous for 5 kb fragment, lanes 5, 6, 10, 12, 14, and 15 show individuals homozygous for the 9 kb fragment, and lanes 1-3, 8 and 9 show the results for heterozygous .individuals.

: Table 2 sets forth the correlation coeffi¬ cients, Rp, calculated as described above, with increases in the measured quantities, for the renin polymorphisms.

Table 2

Blood Pressure .

Systolic Diastolic PRA ^• UA U(Na)V U(K)V U(Cr)V

Bgll 0.08 0.106 -0.185 0.109 0.177 -0.056 0.34

Bglll 0.2 0.135 -0.387 -0.38 0.07 0.04 -0.08

Hindi I I 0.016 -0.336 -0.162 0.137 -0.376 -0.425 -0.2

TaqI 0.224 0.483 -0.373 -0.19 -0.006 -0.057 -0.111

Both the Bglll and TaqI polymorphisms show some correlation with hypertension. They also show high

10 negative correlation with PRA and UA. The Bgll and

Hindlll polymorphisms do not correlate with plasma renin activity; however the Hindlll polymorphism correlates with high diastolic blood pressure and with decreased urinary sodium and potassium, and therefore may be a

15 hypertension predictor, though not specifically of high renin hypertension. The Bgll polymorphism correlates with increased creatinine excretion and may thus predict imbalances in creatinine metabolism.

20 Insulin

• The L/M/U polymorphism and procedures for its detection in the insulin gene are well known and need not be detailed here. This procedure is described, for example, in Ullrich, A. et al, Nucl Acids Res (1982) _c 10:2225-2240. - ~ —

A high correlation with high renin was found for the presence of the U allele. These results are shown, again in terms of correlation coefficients, in

Table 3.

30

Table 3

Rp

(with p value Var:iation

Parameter increase) (t test) predic :ted observed

PRA 0.35 0.03 T T

U(Na)V -0.42 0.03 + +

U(K)V 0.36 0.02 + T

UA 0.31 0.04 T +

Patients with high renin hypertension will show increases in blood pressure, plasma renin activity, and urinary aldosterone and potassium; they will show a decrease in urinary sodium. Therefore, polymorphisms diagnostic for high renin hypertension should show posi¬ tive Rp values for blood pressure, PRA, UA, and U(K)V, and a negative R _p value for U(Na)V. All of the measures associated with high renin hypertension are thus cor¬ rectly correlated with the presence of the U allele.

Other Hypertension Indicators

Polymorphisms in the ANP gene and in the apoAI/apoCIII complex were also shown to correlate with hypertension as measured by blood pressure, although not with the symptomology associated with high renin hyper- tension. Indeed, for these polymorphisms, there was no correlation with high renin levels in plasma. However, because blood pressure itself showed a positive correla¬ tion with these genetic characteristics, they have pre¬ dictive value with respect to hypertension in general.

Several sources for probes useful in detection of polymorphisms of the ANP gene are known in the art. Greenberg, B.D. , et al. Nature (1984) 312:656-658 describe a 2.6 kb genomic clone which encompasses the human preproANP gene. Oikawa, S. , et al, Nature (1984)

309:722-724 describe a human cDNA clone in the Okayama- Berg system encoding an ANP precursor. The probe used herein was designated "clone 6" and was isolated from a human heart cDNA library in λgtlO by probing with the rat ANP-encoding cDNA of Yamanaka, et al, (Nature (1984) 309:719-722) . The nucleotide sequence of the clone 6 probe is shown in Figure 3.

Several polymorphisms of the ANP gene are found as shown in Table 4.

Table 4

More Less

Poly Common Common morphism Probe Enzyme Fraqment Fraqment Frequency

Banll clone 6 Banll 1.8 1.4 0.2

Bgll clone 6 Bgll 6.2 4.1 0.15

Bglll clone 6 Bglll 6.5 9.1 0.06

EcoRI clone 6 EcoRI 11.2 5.2 0.05

BsmI clone 6 BsmI 2.4 1.9; 0.29 0.5

*BstXI clone ^■ 6 BstXI — 0.35 0.1

Seal clone 6 Seal 2.0,8.0 10.0 0.15+0.05

Nsil clone 6 Nsil 7 11 0.025+0.005

*Digestion with BstXI results also in invariant bands of 0.40, 1.20, and.1.90 kb-.

Of these polymorphisms, only the Bgll and Bglll polymorphisms show correlation with high blood pressure. Table 5- shows the correlation coefficients for these two polymorphisms; the table also shows the level of correlation with various parameters associated with high renin hypertension along with the p value associated with the Bglll values.

Table 5

Blood Pressure

Systolic Diastolic PRA UA U(Na)V U(K)V U(Cr)V

Bgll 0.515 0.16 0.02 0.36 0.18 0.42 0.44

Bglll 0.44 0.41 -0.03 -0.02 0.014 0.05 0.26 p value 0.01 0.04 0.7 0.02 0.04 0.015 0.017

'or Bgll

10

There seems to be no correlation with high plasma renin levels. The Bgll polymorphism also correlates with

U(K)V and U(Cr)V, as well as UA.

Correlations have also been found between high

-, c blood pressure and two polymorphisms in the ^* apoAI/apoCIII complex. The nature of these polymorphisms is explained in detail in EPO Publication No. 861905633.4, now published, and incorporated herein by reference. The referenced application describes the

20 detection of a deletion polymorphism in the apoAI gene which correlates to atherosclerosis and also sets forth procedures for its detection. Probes described therein include the apoAI or p5'AI probes; restriction enzymes useful to detect this polymorphism include XmnI, Apal,

2 ₅ Bgll, BstEII, Mspl, PvuII, Rsal, StuI, TaqI, Bglll,

Hindlll, Kpnl, Mstll, Ncol,- and Sspl. Also described in this application is the PvuII polymorphism in the apoCIII gene which represents a G • _* - C conversion about 160 bp from the transcription start of this gene. Pro-

20 cedures for detection of this polymorphism are also set forth. The probe is designated apoCIII, and the restric¬ tion enzyme is PvuII.

Correlation coefficients for hypertension as measured by systolic and diastolic blood pressure.

respectively, are 0.007 and 0.57 for the PvuII polymorphism of the apoCIII gene and -0.88 and -0.14 respectively for the deletion polymorphism of the apoAI gene. Other parameters associated with high renin type hypertension do not show correlations with these polymorphisms, although the U(K)V parameter shows a cor¬ relation coefficient of 0.57 with the apoAI deletion. Thus it appears that these mutations are associated with symptomologies related to clinical indications of hypertension.

In addition, the PvuII polymorphism of the apoCIII gene shows the predicted correlation with aver¬ age diastolic blood pressure for groups of tested patients which would be consistent with an association of that polymorphism with true correlation to high read¬ ings. Thus, for patients who are homozygous for lacking the polymorphism, a sample of fifteen patients showed an average diastolic blood pressure of. 88 ;+ 1.5; twelve patients who were heterozygous for the polymorphism showed an average blood pressure of 96 __ 3.0; a sample of six patients homozygous for the polymorphism had an average blood-pressure of 105 + 2.5. i The estrogen receptor gene shows several polymorphisms which may be correlated to hypertension, as well as to osteoporosis. The gene is described in a paper by Walter, P., et al, Proc Natl Acad Sci USA (1985) 81:7889-7893, incorporated herein by reference. Among the cDNA clones isolated by oligonucleotide hybridization was a 2.1 kb cDNA clone designated OR8. The OR8 clone was converted to estrogen receptor probe (ER probe) by excising a 1.9 kb fragment using TTH111I which cuts at the 5' end and SstI which cuts at the 3' end. The ER probe was then labeled as described above

and used for the detection of polymorphisms in the estrogen receptor gene.

Four polymorphisms were found as set forth in Table 6.

Table : 6

More Less Common Common

Poly¬ Fragment Fragment morphism Probe Enzyme (kb) (kb) Freer.

StuI ER StuI none 7.4 0.05

Mspl 11 Mspl 0.28 0.21 0.45

PssI II PssI none 1.66 0.05

Nsil It Nsil none 3.55 0.08

As seen from Table 6, three of the polymorphisms" occur.at relatively low frequency, while the Mspl polymorphism is present in- almost half of the population.

Co ^' rrelation of the foregoing polymorphisms with the incidence of hypertension has not as yet been verified; however, because of the relevance of estrogen activity to osteoporosis and hypertension, at least some of the foregoing polymorphisms are expected to correlate with the .incidence of these symptomologies.

Polymorphic sites were also found in the Na,K-ATPase gene region. The probe was a 2.0 kb cDNA obtained from a human kidney λgtlO library using probes designed from the disclosed nucleotide sequence for the Na,K-TAPase β subunit set forth in Kawakami, K. , et al, Nucleic Acids Res (1986) 14.:2834-2844, incorporated herein by reference. This 2 kb fragment spans tghe com¬ plete coding sequence for the β subunit and is

designated the "ATPase-β" probe; the 2-5 kb insert obtained as an insert in pHNKB6 of Kawakami could also be used.

Using the ATPase-β probe, two Hindlll and two Mspl polymorphic sites were detected, as shown in Table 7.

Table 7

- More Less Common Common

Poly¬ Fragment Fragment morphism Probe Enzyme (kb) (kb) Frequency

Hindlll-A ATPaseβ Hindlll 11.2 7.4 0.28+0.03

Hindlll-B AtPaseβ Hindlll 4.8 3.7 0.36+0.03

Mspl-A AtPaseβ Mspl 5.9 4.6, 0.35+0.1 1.3

Mspl-B AtPaseβ Mspl 3.5 4.3 0.35+0.1

Hindlll digestion gives background bands at 7.6, 2.5 and 1.3 kb; Mspl digestion at 0.8 and 0.9 kb.

Though correlation of these polymorphisms with hypertension has. not yet been verified, the importance of Na,K-ATPase to the Na ⁺ pump indicates their relevance to this disease. Probes for the Na,K-ATPase α-subunit have also been obtained from the same λgtlO library using probes designed from the complete DNA sequence encoding this subunit as disclosed by Kawakami, K. et al, J Biochem (1986) 100:389-397.

Concluding Remarks

Several of the approximately 10 million polymorphisms existent in the human genome have been shown to be useful predictors of individuals at risk for hypertension. These polymorphisms are detectable, as.

fragments of predictable size obtained by digestion of the genomic DNA of the subject individual with a speci¬ fied restriction enzyme and probing with specific DNA sequences herein described. The availability of this tool for early diagnosis of individuals at risk for hypertension permits the early application of therapeu¬ tic measures to prevent the fatal symptomology of the disease.

In addition to the predictive value for some of the polymorphisms disclosed herein in relation to hypertension, and to certain other parameters, the pat¬ tern of genetic polymorphisms for and individual can be useful in identification and in tracing familial rela¬ tionships, as well as providing a basis for ascertaining genetic patterns associated with other diseases. Thus, the invention ^* also relates to methods to obtain a genetic pattern for a particular individual or for a family by assessment ^" of the presence or absence of one or more of these polymorp isms.