Patent Application
20060166221
This invention relates to diagnosis of essential thrombocythemia. More specifically, this invention relates to diagnosis of essential thrombocythemia by detecting changes in the gene expression or protein levels of type 3 and/or type 12 17β-hyrdroxysteroid dehydrogenases in the platelets of a test subject.
Circulating blood platelets are anucleate, although they retain small amounts of megakaryocyte-derived mRNAs and a fully functional protein biosynthetic capacity2. Essential thrombocythemia (ET) represents a myeloproliferative disorder subtype, characterized by increased proliferation of megakaryocytes, elevated numbers of circulating platelets, and considerable thrombohemorrhagic events, not infrequently neurological3. ET is seen with equal frequency in males and females, although an additional female incidence peak at age 30 may explain the apparent higher disease prevalence in females. The molecular basis of ET remains to be established, although historically it has been considered a “clonal” disorder4. Causative mutations have been identified in the thrombopoietin gene. However, these mutations appear to be uncommon and restricted to rare individuals with familial thrombocythemia5. Other than the exaggerated platelet volume evident in subsets of ET platelets, the cells remain morphologically indistinguishable from their normal counterparts. No functional or diagnostic test is currently available for ET, and it remains to be diagnosed by exclusion.
The use of gene expression profiling in molecular classification of human cancer is well-documented. The feasibility of platelet profiling using apheresis techniques has also been demonstrated2,6. However, it remains to be established whether gene expression profiling can be applied to diagnosis of poorly-understood myeloproliferative disorders, such as ET.
17β-hyrdroxysteroid dehydrogenases (17β-HSD) function in the formation and inactivation of all active androgens and estrogens, with substrate interconversion regulated by the oxidative state of the NADP/NAD(P)H cofactors9. To date, gene products encoding twelve types of 17β-HSD enzymes have been described, although the type 6 and 9 genes have been only characterized in rodents14. The HSD type 3 enzyme is generally regarded as testes-specific, although rare SAGE tags have been identified in CGAP tissues including brain (7×10−6), skeletal muscle (9.3×10−6), and prostate (1.6×10−5)10. The 17β-HSD3 enzyme specifically mediates the catalytic interconversion involving 4-androstenedione and testosterone. Molecular defects of the HSD17B3 gene are causally implicated in male pseudohermaphroditism11. While steroidogenic pathways are incompletely characterized in platelets, it has been demonstrated that megakaryocytes (Mk) express the glucocorticoid receptor, and that both Mk and platelets selectively express estrogen receptor (ER) β and androgen receptor mRNA and protein, to the exclusion of ER α or progesterone receptor12. Furthermore, Mk express functional 3β-HSD, known to catalyze the essential step in the transformation of 5-pregnen-3β-ol and 5-androsten-3β-ol steroids into the corresponding Δ4-3-keto-steroids, i.e., progesterone as well as the precursors of all androgens, estrogens, glucocorticoids and mineralocorticoids. Indeed, Mk-derived estradiol triggers megakaryocyte proplatelet formation in vitro, an effect that is blocked by inhibition of 3β-HSD activity13.
The present invention provides for the first time evidence showing that distinct subtypes of the steroidogenic 17β-HSDs are functionally expressed in human blood platelets, and that the expression patterns of HSD17B3 and HSD17B12 are distinctly associated with ET manifest by quantitative and qualitative platelet defects. Therefore, the present invention has identified the first diagnostic molecular signature for ET.
The present invention provides methods of diagnosing ET based on detecting the levels of gene expression or the levels of the proteins of type 3 and type 12 17β-hyrdroxysteroid dehydrogenases in a test subject. Test subjects include any individuals of any age, more preferably to individuals having an abnormally high platelet count.
For the purpose of detecting gene expression or protein levels of type 3 and type 12 17β-hyrdroxysteroid dehydrogenases, a blood sample is taken from a test subject. Although not absolutely necessary, it is preferred that platelets are isolated from the blood sample for use in an assay. More preferably, isolated platelets are lysed in order to obtain mRNAs or cell lysates for use in the assay.
In one embodiment, diagnosis of ET is based on detecting a lower level of expression of the HSD17B3 gene in a subject, as compared to a control level (i.e., a level of gene expression in normal individuals without ET). Preferably, diagnosis is based on a level of HSD17B3 expression that is lower than a control level by at least 2-3 fold; more preferably, at least 4-6 fold.
In another embodiment, diagnosis of ET is based on detecting a higher level of expression of the HSD17B12 gene in a subject, as compared to a control level. Preferably, diagnosis is based on a level of HSD17B12 expression that is higher than a control level by at least 4-5 fold; more preferably, by at least 10-12 fold, or even 20-30 fold.
In yet another embodiment, diagnosis of ET is based on detecting a lower level of expression of the HSD17B3 gene, as well as a higher level of expression of the HSD17B12 gene in a subject, as compared to control levels of expression of the respective genes.
In still another embodiment, diagnosis of ET is based on determining the ratio of HSD17B12 gene expression versus HSD17B3 gene expression, and comparing the ratio with that obtained from normal subjects. An increased ratio is indicative of ET in a test subject. Preferably, the diagnosis is based on an increase in the HSD17B12:HSD17B3 ratio by at least 20 fold; even more preferably at least 50 fold; and most preferably, 75 to 100 fold.
In another embodiment, diagnosis of ET is based on determining the log2 ratio of HSD17B12 gene expression versus HSD17B3 gene expression. Generally speaking, a log2 ratio greater than 1 is indicative of ET in the test subject. Preferably, diagnosis is based on a log2 ratio greater than 2, even more preferably, greater than 4.
In a further aspect of the invention, diagnosis of ET is based on detecting a lower level of the 17β-HSD3 protein, and/or a higher level of the 17β-HSD12 protein in a test subject, as compared to control protein levels. Similarly, diagnosis of ET can be based on detecting an increase in the ratio of the level of 17β-HSD12 versus the level of 17β-HSD3 in a subject, as compared to a control ratio.
The present inventors have discovered that distinct subtypes of steroidogenic 17β-hyrdroxysteroid dehydrogenases (17β-HSDs) are functionally present in human blood platelets, and that the expression patterns of HSD17B3 and HSD17B12 are distinctly associated with essential thrombocythemia (ET), a myeloproliferative disorder characterized by increased proliferation of megakaryocytes and elevated numbers of circulating platelets. More specifically, it has been determined that the gene (HSD17B3) encoding type 3 17β-hydroxysteroid dehydrogenase is selectively down-regulated in ET platelets, with an induction of gene expression for the type 12 enzyme. Accordingly, the present invention provides diagnostic assays for ET based on detecting the levels of gene expression or the levels of the proteins of type 3 and type 12 17β-hyrdroxysteroid dehydrogenases in a subject.
As used herein, the abbreviation “17β-HSD” represents the protein form of steroidogenic 17β-hyrdroxysteroid dehydrogenase, and “HSD17B” (in italics) represents the corresponding nucleic acid encoding the 17β-hyrdroxysteroid dehydrogenase.
The present methods for diagnosing ET can be applied to any individuals of any age, more preferably to individuals having an abnormally high platelet count. Normal platelet counts generally range from 150,000/microliter of blood to 350,000/microliter of blood.
According to the present invention, the diagnostic methods can be either nucleic acid-based assays or protein-based assays. That is, the methods can be based on detecting the level of expression of the relevant gene, or based on detecting the level of the expressed protein product, in a sample taken from a test subject containing platelet.
Typically, a blood sample is taken from a test subject. The amount of blood taken may vary depending upon the particular assay to be used. Generally speaking, for nucleic acid-based assays, about 0.1 ml to 15 ml, or preferably about 1 ml to 10 ml of blood is taken from a subject. For protein-based assays, about 1 ml to 20 ml, or preferably about 5-10 ml, of blood is taken from a subject.
Although whole blood samples can be used directly in the assays as described herein, preferably, platelets are isolated from whole blood samples for use in the assays. Isolation of platelets can be achieved by standard techniques known in the art, e.g., apheresis, gel filtration and leukocyte immunodepletion by CD45-coupled magnetic micrbeads. More preferably, isolated platelets are further processed, e.g., lysed, in order to obtain mRNAs or cell lysates for use in the assays. The levels of mRNA of a gene of interest can be determined by a variety of assays, such as Northern Blot, RT-PCR, among others. The levels of a protein of interest can also be determined by using antibodies specific for the protein in standard assays, such as Western Blot, ELISA, among others.
In one embodiment, the present invention provides a method of diagnosing ET in a subject by detecting a lower level of expression of the HSD17B3 gene in the subject, as compared to levels of expression of the gene in normal individuals.
Levels of expression of the HSD17B3 gene in normal individuals, i.e., individuals without ET, can be determined in assays run side-by-side with the assay on a test subject. Alternatively, gene expression levels in normal individuals can be predetermined or established. A single value, such as the median or mean expression level, can be determined from a group or population of normal individuals for use as control in the diagnosis.
As described hereinabove, the level of expression of a gene, which is reflected by the level of the relevant mRNA, can be determined by a variety of assays, including Northern Blot analysis, RT-PCR, and the like.
According to the present invention, the diagnosis of ET is based on a level of HSD17B3 expression that is lower than a control level, preferably by at least 2-3 fold. More preferably, the diagnosis is based on a level of HSD17B3 expression that is lower than a control level by at least 4-6 fold.
In another embodiment, the present invention provides a method of diagnosing ET in a subject by detecting a higher level of expression of the HSD17B12 gene in the subject, as compared to levels of expression of the gene in normal subjects.
According to the present invention, the diagnosis of ET is based on a level of HSD17B12 expression that is higher than a control level, preferably by at least 4-5 fold. More preferably, the diagnosis is based on a level of HSD17B12 expression that is higher than a control level by at least 10-12 fold, or even 20-30 fold.
In yet another embodiment, the present invention provides a method of diagnosing ET in a subject by detecting a lower level of expression of the HSD17B3 gene, as well as a higher level of expression of the HSD17B12 gene in the subject, as compared to levels of expression of the genes in normal subjects.
In still another embodiment, the present invention provides a method of diagnosing ET in a subject by detecting the levels of expression of the HSD17B3 gene and the HSD17B12 gene in the subject, determining the ratio of HSD17B12 gene expression versus HSD17B3 gene expression, and comparing the ratio with that obtained from normal subjects. An increased ratio is indicative of ET in the test subject.
According to the present invention, diagnosis of ET based on the ratio of HSD17B12: HSD17B3 gene expression is more sensitive than diagnosis based on the change of expression of either gene alone. Preferably, diagnosis of ET is based on an increase in the HSD17B12: HSD17B3 ratio by at least 20 fold; and more preferably, at least 50 fold; and even more preferably, at least 75 to 100 fold.
In another embodiment, diagnosis of ET is based on determining the log2 ratio (i.e., the log2 value of the ratio) of HSD17B12 gene expression versus HSD17B3 gene expression. Generally speaking, a log2 ratio greater than 1 is indicative of ET in the test subject. Preferably, diagnosis is based on a log2 ratio greater than 2, even more preferably, greater than 4.
In a further aspect, the present invention provides a method of diagnosing ET in a subject by detecting a lower level of the 17β-HSD3 protein in the subject, as compared to levels of the 17β-HSD3 protein in normal individuals. As described hereinabove, the level of a protein can be determined using an antibody specific for the protein by a variety of assays, including Western Blot analysis, ELISA, among others.
According to the present invention, the diagnosis of ET is based on a level of the 17β-HSD3 protein that is lower than a control level determined from normal individuals, preferably by at least 2-3 fold. More preferably, the diagnosis is based on a level of the 17β-HSD3 protein that is lower than a control level by at least 4-6 fold.
In another embodiment, the present invention provides a method of diagnosing ET in a subject by detecting a higher level of the 17β-HSD12 protein in the subject, as compared to levels of the 17β-HSD12 protein in normal subjects.
The diagnosis of ET is based on a level of the 17β-HSD12 protein that is higher than a control level, preferably by at least 4-5 fold. More preferably, the diagnosis is based on a level of the 17β-HSD12 protein that is higher than a control level by at least 10-12 fold, or even 20-30 fold.
In yet another embodiment, the present invention provides a method of diagnosing ET in a subject by detecting a lower level of the 17β-HSD3 protein, as well as a higher level of the 17β-HSD12 protein in the subject, as compared to levels of the proteins in normal subjects.
In still another embodiment, the present invention provides a method of diagnosing ET in a subject by detecting the levels of the 17β-HSD3 protein and the 17β-HSD12 in the subject, determining the ratio of the two proteins (17β-HSD12:17β-HSD3), and determining an increase of the ratio when comparing to normal subjects as indicative of ET.
Alternatively, diagnosis of ET in a subject is based on determining the Log2 ratio of the two proteins (17β-HSD12:17β-HSD3). A Log2 ratio greater than 1 is indicative of ET in the test subject. Preferably, diagnosis of ET is based on a Log2 ratio greater than 2; more preferably, by a ratio greater than 4.
The present invention is further illustrated by the following non-limiting example.
Patient Selection and Characterization
Patients were enrolled from the larger pool of patients referred to the Division of Hematology for evaluation of thrombocytosis. All patients provided informed consent for an IRB (Institutional Review Board)-approved protocol completed in conjunction with the Stony Brook University Hospital General Clinical Research Center. Standard hematological criteria were followed for the diagnosis of essential thrombocythemia, reactive thrombocytosis, and other myeloproliferative disorders24,25. Both sex- and age-distribution paralleled prevalence figures for ET, with a M:F ratio of 1:2.3, and age at diagnosis ranging from 23 to 78 years old. Platelet counts at the time of blood isolation ranged from normal (reflecting treatment) to 1,724,000/μL. Patient utilization of platelet-lowering drugs (i.e., hydroxyurea, analgrelide, or untreated) was recorded at the time of platelet isolation and purification. Detailed patient characteristics are set forth in Table 1.
Platelet Molecular Studies
Platelets were obtained by apheresis or from peripheral blood (10 mL), and were isolated essentially as previously described, utilizing gel-filtration and CD45-coupled magnetic micro-beads for leukocyte immunodepletion2. The final platelet-enriched product contained no more than 3-5 leukocytes per 1×105 platelets. Peripheral blood leukocytes from three healthy donors were isolated as previously described2. Pure cellular pellets were resuspended in 10 mL of TRIZOL reagent (Invitrogen, Carlsbad, Calif.), transferred into DEPC (diethylpyrocarbonate)-treated Corex tubes, and serially purified and precipitated using isopropanol26. Platelet mRNA quantification and integrity were established using an Agilent 2100 Bioanalyzer, and quantitative reverse transcription(RT)-PCR was performed using fluorescence-based real-time PCR technology (TaqMan Real-Time PCR, Applied Biosystems, Foster City, Calif.). Oligonucleotide primer pairs were generated using Primer3 software, designed to generate PCR products of approximately 200-bp at the same annealing temperature (71° C.). HSD17B3-specific primers were: forward (5′AAATGTGATAACCAAGACTGC 3′ [bp 755-775] (SEQ ID NO: 1)); reverse (5′CTTGGTGTTGAGCTTCAGGTA 3′ [bp 956-936] (SEQ ID NO: 2)); HSD17B12-specific primers were: forward (5′ TGAATACTTTTGGATGTTCCTGA 3′ [bp 496-519] (SEQ ID NO: 3)); reverse (5′ AGTCTTGGTTGCAGAATAGATGGT 3′ [bp 634-611] (SEQ ID NO: 4)); HSD17B11-specific primers were: forward (5′ TGGATATAAAATGAAAGCGCAATA 3′ [bp 1067-1090] (SEQ ID NO: 5)); reverse (5′ ATCAGCTTTTGGCTAAAGAACAAG 3′ [bp 1265-1242] (SEQ ID NO: 6)); F7-specific primers were: forward (5′ TCCTGTTGTTGGTGAATGG 3′ [bp 734-753] (SEQ ID NO: 7)); reverse (5′ GTACGTGCTGGGGATGATG 3′ [bp 933-915] (SEQ ID NO: 8)); β-actin-specific primers were as previously described2. Purified platelet mRNA (4 μg) was used for first strand cDNA synthesis using random hexamers and SuperScript II reverse transcriptase (Invitrogen, Carlsbad, Calif.). For RT-PCR analysis, the RT reaction was equally divided among primer pairs and used in a 40-cycle PCR reaction for each target gene using two-step cycles of 94° C. for 20 seconds and 71° C. for 1 min. mRNA levels were quantified by monitoring real-time fluorimetric intensity of SYBR green I. Relative mRNA abundance was determined from triplicate assays performed in parallel for each primer pair, and calculated as previously described27,28. For some patients, high molecular-weight genomic DNA was isolated from peripheral blood leukocytes for PCR-based amplification and sequencing of exon-intron boundaries29.
Gene expression profiles were completed using the 22, 283 HU133A probe set (Affymetrix, Santa Clara, Calif.). Total cellular RNA (5.8 μg) was used for cDNA synthesis using SuperScript Choice system (Life Technologies, Rockville, Md.) and an oligo(dT) primer containing the T7 polymerase recognition sequence, followed by cDNA purification using phenol/chloroform extraction and ethanol precipitation. In vitro transcription was completed in the presence of biotinylated ribonucleotides using a BioArray HighYield RNA Transcript Labeling Kit (Enzo Diagnostics, Farmingdale, N.Y.). After metal-induced fragmentation, 10 μg of the biotinylated cRNA was hybridized to the oligonucleotide probe array for 16 hours at 45° C. After washing, the cRNA was detected with streptavidin-phycoerythrin (Molecular Probes, Eugene, Oreg.), and analysis was completed using a Hewlitt-Packard Gene Array Scanner. The fluorescence intensity of each probe was quantified using Affymetrix GeneChip software (MAS version 5.0), calculated as an average difference for each gene set obtained from 16 to 20 paired (perfectly matched and single nucleotide-mismatched) 25-bp oligonucleotides. The software was designed to exclude “positive calls” in the presence of high average differences with associated high mismatch intensities.
Bioinformatic and Statistical Analyses
Microarray data were visualized and analyzed using GeneSpring (version 7.0) software (Agilent, Palo Alto, Calif.). Data were normalized by dividing each measurement by the 50th percentile of all measurements in that sample, and each gene was divided by the median of its measurements in all samples. Normalized median ratios of individual genes were log2-transformed, filtered for presence across arrays, and selected for expression levels as detailed. Prior to unsupervised hierarchical clustering of the uncentered Pearson correlation similarity matrix, the data were filtered for gene expression across phenotypic cohorts, defined as those genes present or marginal in a minimum of 80% of platelet samples (yielding 2,906 transcripts). A subset of genes was culled from the 2,906-gene list to specifically delineate those transcripts uniquely expressed in platelets; this platelet-restricted subset (N=126) was delineated by removing genes expressed in 3/3 leukocyte microarrays. A non-parametric analysis of variance test (ANOVA) was performed to identify differentially expressed genes using the Benjamini and Hochberg method to lower the false discovery rate (p<0.01). All statistical analyses were completed using SPSS (Statistical Package for Social Sciences, version 11.5) software.
Functional 17β-HSD Studies
Functional studies for platelet 17β-HSD3 activity were completed using gel-filtered platelets (GFP). Briefly, 1.5×108 platelets were solubilized and freeze-thawed in HSD buffer containing (20 mM Tris-HCl (pH 8.0), 5 mM MgCl2, 50 mM NaCl, 5% Glycerol, 10 mM DTT, 1.5 mM NAD (D-5755, Sigma Co., St. Louis, Mo.)). The reaction was started by addition of 1 μL (12.7 pM) of [1,2,6,7-3H(N)]-testosterone (specific activity 78.5 Ci/mM) (Dupont/NEN) to equivalent protein aliquots. A 30- or 90-minute reaction was allowed to proceed at 25° C., and quenched at −20° C. in the presence of cold testosterone (1 mg/mL) and androstenedione (1 mg/mL). Steroid extraction was performed twice using 150 mL of ethyl acetate. Both fractions were combined and air-dried under vacuum. Androstenedione and testosterone were then separated by thin layer chromatography using 4:1 (v/v) chloroform:ethyl acetate. Discrete steroid fractions were visualized and extracted from silica gels using ethanol, and quantified by liquid scintillography. Purified hydroxysteroid dehydrogenase from Pseudomonas testosteroni (Sigma Co., St. Louis, Mo.) diluted to 10 mg/mL in HSD buffer served as control for some experiments. Mouse testis extract prepared from a C57/B16 mouse served as standard for platelet 17β-HSD3 quantification, and was prepared by homogenization in 3 mL of HSD buffer. After centrifugation at 5000 g for 5 minutes, supernatants were gel-filtered in HSD-equilibrated Centrisep spin columns prior to use. Protein quantification of all samples was determined by Lowry method as previously described30.
1Age at diagnosis
2Sample source: B - peripheral blood; P—plateletpheresis
3Platelet count at time of sample collection (normal range 150-350); note that for patients with normal platelet counts at time of blood isolation, the platelet counts in parentheses are highest pre-treatment determinations
4Refers to treatment at the time of platelet isolation; A—Anagrelide; H—Hydroxyurea; N—Not treated
5M—Microarray; P - Quantitative RT-PCR
6Originally given the diagnosis of polycythemia rubra vera (PV)
7Secondary (post-splenectomy) thrombocytosis
The genetic profiles of highly-purified apheresis platelets isolated from 6 ET patients (4 females, 2 males) and 5 normal, healthy controls demonstrated distinctly different molecular signatures (
An unsupervised, hierarchical clustering algorithm was used to group normal and ET platelet genes on the basis of similarities of gene expression (
Several methods of computational analyses were completed to identify genes that could discriminate between ET and normal platelets. An initial one-way ANOVA analysis identified 170 genes that were differentially expressed, the majority of which (141) were up-regulated in ET platelets, with only 29 genes down-regulated in ET platelets compared to normal platelets. Functional cluster analysis of this limited set of differentially-expressed genes (
Because of inherent difficulties in analysis of microarray datasets, the data were re-analyzed by computing t-statistics of ET versus normal platelets for each gene. The t-values were ranked by absolute magnitude (thereby incorporating inter-sample variability in expression ratios), and were then ranked by the magnitude of the test statistic numerator, a measure of the biological difference in expression ratios. By applying a rigid 5-fold difference in pair-wise expression as the cut-off, 163 genes were identified that were up-regulated in ET. Only a small number (5) of genes were down-regulated (
A single platelet-restricted gene HSD17B3 (encoding the type 3 17β-hyrdroxysteroid dehydrogenase [17β-HSD3]) was expressed in all normal arrays, and uniquely underexpressed in ET platelets compared to normal platelets (
To validate and extend these results, a quantitative RT-PCR (qRT-PCR) assay was developed and was applied to the original ET cohort and an expanded cohort of normal controls, specifically collected to exclude potential gender-bias in HSD17B gene expression. The results paralleled those found by microarray, demonstrating approximately 4.5-fold greater HSD17B3 transcript (compared to HSD17B12) in normal platelets (p≦0.001) and concomitant approximately 27-fold greater HSD17B12 transcript expression (compared to HSD17B3) in ET platelets (p≦0.03). The reciprocal changes HSD17B3 and HSD17B12 amount to an aggregate about 2-log change in intracellular HSD17B3:HSD17B12 transcript levels between normal and ET platelets (
Because the qRT-PCR data established an absolute decrease in HSD17B3 transcript level in ET platelets, preliminary genomic analyses of HSD17B3 were completed in 4 of the 6 ET patients. The 11 exons and intron-exon boundaries were amplified and sequenced, with the identification of a single heterozygous A insertion (not involving the splice junction site15) in the first intron of one patient (ET1). Thus, there was no evidence that a small deletion or missense mutation affecting HSD17B3 transcript stability was causally implicated in the pathogenesis of HSD17B3 transcript loss.
The distinct patterns of HSD17B expression identified by microarray and confirmed by qRT-PCR were then extended to a larger cohort of 20 ET patients (6 original ET patients and 14 newly-studied individuals), now uniformly analyzed using peripheral blood as the starting source for platelet analysis (unlike the cumbersome apheresis technique used in the original cohorts). The qRT-PCR results were entirely concordant for all individuals studied, demonstrating that 17BHSD12:17BHSD3 transcript ratios reliably predicted the ET phenotype in all patients studied to date (p<0.0001) (
To confirm that platelets retained functional 17β-HSD3 activity (and to compare this activity between normal and ET platelets), the oxidative conversion of testosterone to 4-androstenedione was quantified. Entirely consistent with the genetic data, normal platelets retained 17β-HSD3 activity, providing for the first time evidence that non-testicular sources retain functional capacity in the penultimate step of androgen biosynthesis. Furthermore, the platelet-derived 17β-HSD3 activity was not inconsequential, providing nearly 10% of the capacity found in testis (
17. Wattel, E. et al. Androgen therapy in myelodysplastic syndromes with thrombocytopenia: a report on 20 cases. Br J Haematol 87, 205-208 (1994).
The work described in this invention was supported in part from grants from the U.S. National Institutes of Health, NHLBI (HL04239, HL76457 and HL49141). The government may have rights in this invention.
