Demonstration and characterization of specific binding sites for factor VIII/von Willebrand factor on human platelets.

The presence of specific Factor VIII/von Willebrand factor (FVIII/vWF) binding sites on human platelets has been demonstrated by using 125I-FVIII/vWF and washed human platelets. Binding is ristocetin-dependent and increases in proportion to the concentration of ristocetin from 0.2 to 1 mg/ml. Binding of 125I-FVIII/vWF to platelets can be competitively inhibited by unlabeled human or bovine FVIII/vWF, but not by human thrombin, fibrinogen, alpha 2-macroglobulin, equine collagen, or a lectin of Ricinus communis. Scatchard analysis of binding data indicated that the dissociation constant of FVIII/vWF receptors is 0.45--0.5 nM. There are 31,000 binding sites per platelet at 1 mg/ml of ristocetin concentration. The optimal pH range for binding is from 7.0 to 7.5. At a concentration of 2 mM, EGTA inhibits 86% of the binding; however, 20 mM of Ca++, Mg++, or EDTA have little effect. Binding sites for FVIII/vWF were found only on platelets, and no significant binding was detected with human erythrocytes or polymorphonuclear leukocytes.

Inhibition of tumor cell-induced platelet aggregation and experimental tumor metastasis by the synthetic Gly-Arg-Gly-Asp-Ser peptide.

The mechanism by which the murine fibrosarcoma clone PAK 17.15 induces platelet aggregation [tumor cell-induced platelet aggregation (TCIPA)] was studied because platelet activation by this clone is necessary for metastasis to the lungs. PAK 17.15 TCIPA was completely inhibited by ADP-clearing enzymes, such as apyrase, or a mixture of creatine phosphate and creatine phosphokinase. Thrombin and collagen were not involved in PAK 17.15 TCIPA. Further studies showed that ADP is most likely secreted from activated platelets and that membrane protein(s) on PAK 17.15 cells are responsible for platelet activation. Inasmuch as ADP-dependent platelet aggregation requires fibrinogen and can be inhibited by the Gly-Arg-Gly-Asp-Ser (GRGDS) synthetic peptide, the effect of this peptide on PAK 17.15 TCIPA was studied. PAK 17.15 TCIPA was completely inhibited by the GRGDS peptide (0.4 mM) but not by a control peptide, Gly-Arg-Gly-Glu-Ser (0.8 mM). In addition, the GRGDS peptide inhibited adhesion of PAK 17.15 cells to immobilized fibronectin. As expected, the GRGDS peptide almost completely inhibited lung colonization by iv injected PAK 17.15 cells in C57BL/6 mice. Our results indicate that GRGDS may inhibit pulmonary metastases by interfering with TCIPA as well as with tumor cell adhesion to extra-cellular matrix components in the host.

Functional role of platelets in experimental metastasis studied with cloned murine fibrosarcoma cell variants.

The involvement of platelets in experimental metastasis was studied with cloned cell lines derived from PAK 17, a recently induced methylcholanthrene-induced C57BL/6 mouse fibrosarcoma. Tumor cell-induced platelet aggregation and lung colonization assays were used to distinguish three major stable phenotypes among the clones: a low metastatic-low platelet aggregating type, e.g., clone PAK 17.12; a low metastatic-high platelet aggregating type, e.g., clone PAK 17.14; and a high metastatic-high platelet aggregating phenotype, e.g., clone PAK 17.15. Clones with high metastatic but low platelet aggregating potential were not observed in the study. Intravenously injected PAK 17.14 and PAK 17.15 cells, but not PAK 17.12 cells, induced greater than 50% reductions in circulating platelet levels in C57BL/6 mice. Since highly metastatic clone PAK 17.15 cells consistently induced high levels of tumor cell-induced platelet aggregation regardless of the platelet donor, it was selected to study the relationship between its tumor cell-induced platelet aggregation and lung colonizing abilities. (a) A 93% decrease in lung colony number resulted in mice injected with 100 micrograms of prostacyclin immediately before injection of clone PAK 17.15 cells. Prostacyclin was also able to inhibit, in a dose dependent fashion (0-5 ng), platelet aggregation induced by clone PAK 17.15 cells in vitro. (b) A 92% reduction in lung colony number occurred in mice showing marked thrombocytopenia following injection of 100 micrograms of rabbit anti-mouse platelet antibody 24 h before tumor cell injection. (c) A greater than 80% reduction in clone PAK 17.15 lung colony number was observed in mice rendered thrombocytopenic by i.v. injection of 0.038 units of neuraminidase 24 h before i.v. injection of 10(5) tumor cells. These results suggest that platelets are required for successful lung colonization by clone PAK 17.15 cells. However, the presence in this fibrosarcoma of high platelet aggregating-poorly metastatic cells, such as clone PAK 17.14, demonstrates that while the ability to aggregate platelets is necessary for successful metastasis by some tumor cells, it is insufficient if tumor cells lack other critical properties required for completion of the metastatic cascade.

The lack of correlation between experimental metastatic potential and platelet aggregating activity of B16 melanoma clones viewed in relation to tumor cell heterogeneity.

Two widely used B16 melanoma cell lines of low and high lung colonizing potential (B16-F1 and B16-F10) were compared in their ability to induce platelet aggregation. The results of these experiments showed a reproducible difference in platelet aggregating activity of these two cell lines which directly correlated with their lung colonizing potentials. However, when clones were derived from these heterogeneous cell lines and tested for experimental metastatic potential, platelet aggregating ability and Met-72 expression, no correlation could be attached to the platelet aggregating activity of the clones. Results of these experiments provide direct evidence that platelet aggregation is not an accurate index of experimental metastatic potential of tumor cell clones, nor is it an essential trait of all metastatic cells. The ability of tumor cells to induce platelet aggregation is examined and discussed in the context of cellular heterogeneity.

Total parenteral nutrition in mice bearing a metastatic carcinoma: tumor growth, metastasis and immunologic parameters.

The role of dietary manipulation of tumor growth, metastasis and immunologic parameters was studied in mice bearing Lewis lung carcinoma. Fourteen days following subcutaneous tumor implant, groups with tumor and their non-tumor bearing counterparts were assigned to one of the following feeding protocols: total parenteral nutrition (TPN), per oral (PO) intake of the parenteral diet, an oral casein diet (CAS), or electrolyte infusion plus the casein diet (ELECT). Intakes of energy and nitrogen were similar among all groups. Mice were killed 12 days later and peritoneal macrophages were tested for phagocytic activity. Tumor growth and metastasis were decreased from both infusion regimens with minimal loss of body weight as compared with casein fed mice. PO mice also showed lower tumor weight but metastasis was as great as in the casein group. Non-tumor-bearing infused mice showed depressed thymic weight, but thymic weight was not further reduced in tumor-bearing infused mice. PO feeding afforded no such protection in the presence of the carcinoma. Splenomegaly was observed in tumor-bearing mice on all regimens, but mice maintained on the parenteral diet demonstrated the largest proportion of macrophages containing nuclear debris. Analysis of free macrophages indicated no effect of diet regimen on non-immune phagocytic activity in both tumor-free and tumor-bearing mice. Possible alteration of splenic macrophage intracellular digestive capacity or phagocytic activity was suggested as a result of TPN.

Genetic selection, general characterization, and histology of atherosclerosis-susceptible and -resistant Japanese quail.

A new animal model of atherosclerosis has been developed through genetic selection of Japanese quail (Coturnix coturnix japonica) into susceptible (SUS) and resistant (RES) lines. Characterization of the selected quail has shown that the RES birds were resistant to the disease and developed little atherosclerosis on a diet containing 1% cholesterol. The SUS birds were sensitive and developed severe atherosclerosis in 8-9 wks on a diet containing only 0.5% cholesterol. The histology of the progression of atherosclerosis in the SUS quail was studied. It bore a morphological similarity to that of human arterial atherosclerosis. The atherosclerotic plaque was characterized by intimal thickening, the presence of foam cells, the proliferation of smooth muscle cells and/or fibroblasts, and the formation of scar with collagen deposition. We believe that these two lines of quail may serve as a valid animal model for the study of the genetic and biochemical basis of cholesterol-induced atherosclerosis.

A combined approach for quantitation of each specific HLA-A or -B antigen expressed on cells.

In order to measure the amount of each individual HLA-A or -B antigen expressed in or on a cell without relying on monoclonal antibodies to different specific HLA antigens, we have developed a combined approach that consists of two separate measurements. The first measurement is to determine the relative quantities of different HLA-A and -B antigens in lysates of whole cells using IEF gel electrophoresis, immunoblotting with 171.4 anti-HLA heavy chain monoclonal antibody, and scanning densitometry. The second measurement is to determine the concentration of total HLA antigens expressed in or on a cell using an enzyme-linked immunoassay or a surface binding assay based on W6/32 anti-HLA monoclonal antibody. The quantity of each specific HLA antigen in or on a cell then is calculated from the results of these two measurements. To validate this combined approach, we conducted studies to show that the amounts of different specific HLA antigens measured by this approach were linearly correlated with those measured by FITC-labeled monoclonal antibodies and immunofluorescence flow cytometry in platelets and lymphoblastoid cell lines. We also demonstrated that the relative quantities of different HLA-A and -B antigens determined from the lysates of whole cells were the same as those expressed on the cell surface. These findings indicated that the newly developed combined approach can be applied to quantify each specific HLA-A or -B antigen expressed in or on a cell.

Measurement of relative quantities of different HLA-A and -B mRNAs in cells by reverse transcription-polymerase chain reaction and denaturing gradient gel electrophoresis.

In order to determine the relative quantities of different HLA-A and -B mRNAs in cells, we have developed a simple and reliable method by using reverse transcription-polymerase chain reaction (RT-PCR), denaturing gradient gel electrophoresis (DGGE) and phosphor imaging analysis. Cytoplasmic RNA from lymphoblastoid cell lines with well-characterized HLA phenotypes are reversely transcribed with a primer specific for all HLA-A and -B antigens. The first-strand cDNA is used as template for quantitative PCR. The primer pair used for quantitative PCR are specific for all class I HLA and one of the primers is labeled with [gamma-32P]ATP. The amplified sequences include parts of exon 2 and exon 3 which contain most polymorphic residues in class I HLA molecules. The RT-PCR products containing the amplified HLA-A and -B sequences are separated by DGGE. The radioactivities of different DNA bands separated in denaturing gradient polyacrylamide gels are measured by phosphor imaging and used to determine the relative amounts of HLA-A and -B mRNAs. This approach is validated by using samples containing known quantities of different HLA-A and -B mRNA transcripts and confirmed by S1 nuclease protection assay. The combined RT-PCR/DGGE approach therefore provides a simple and reliable method for quantitation of relative amounts of different HLA-A and -B mRNAs in cells. This method should also be useful for studying the expression of other highly conserved and duplicated genes.

Enzyme-linked immunoassay for anti-HLA antibodies--an alternative to panel studies by lymphocytotoxicity.

In order to provide a simple and specific assay for the detection and quantitation of IgG and IgM anti-HLA antibodies in sera, HLA antigens purified from a pool of 240 random donor platelets were used to develop a solid-phase enzyme-linked immunoassay (EIA). The reference values for identifying the presence of IgG or IgM anti-HLA antibodies were determined by assaying sera from 39 healthy individuals without prior HLA alloimmunization. The assay was evaluated by studying sera from 122 patients who had been characterized previously for panel reactive antibodies by the lymphocytotoxicity assay (LCA). A significant linear correlation between two assays was noted (r = 0.8, P = 0.0001). Further analyses of the data demonstrated that the newly developed EIA has 100% specificity and 95.3% sensitivity as compared with the LCA. Additional studies revealed that patients whose PRA increased or decreased over time were in parallel with antibody levels measured by EIA. When the EIA was used to measure anti-HLA antibody titers, it was more sensitive than the LCA. Since the EIA is sensitive, specific, and technically less demanding, it should provide an useful alternative to reduce the number of the more laborious panel studies for monitoring anti-HLA antibody status in candidates for organ transplantation and recipients of blood transfusions.

Class I HLA molecules on human erythrocytes. Quantitation and transfusion effects.

HLA class I molecules were quantitated on erythrocytes from individuals expressing either high or low levels of such antigens. Quantitative determinations were accomplished using 125I-labeled Fab fragments of the anti-HLA monoclonal antibody W6/32 in a competitive binding assay. The experimental conditions of the test system were established using red cells from an individual found to express high levels of red cell HLA when examined by flow cytometry. The competitive binding assay met the requirements of ligand specificity and specific binding saturability. Scatchard analysis revealed that there were 1684 +/- 39 (mean +/- SD) HLA molecules/red cell. In two other donors in whom erythrocyte HLA was undetectable by flow cytometry specific binding of the 125I-W6/32 Fab fragments was clearly demonstrated, indicating the presence of HLA on red cells of these donors as well. The number of HLA molecules/red cell was estimated to be between 100 and 200 for these individuals. Thus, in a blood transfusion unit, the number of HLA molecules contributed by the red cells is comparable to that of the leukocytes. Blood highly depleted of leukocytes and platelets and selected from donors with low amounts of red cell HLA was not beneficial (when transfused to selected patients) in that their sensitizing effects were not significantly different from regular blood transfusions. These results show that the amount of HLA antigens on red cells, while low if compared with other cell types, is significant in terms of the absolute antigenic content of blood transfusions. They also show that transfusion of blood units containing HLA antigens in concentrations as low as can be achieved with current technology were not useful in preventing HLA sensitization in patients at risk.

Functional involvement of thrombospondin in platelet aggregation induced by low versus high concentrations of thrombin.

Thrombospondin (TSP) is a major platelet secretory glycoprotein. Earlier studies of various investigators demonstrated that TSP is the endogenous platelet lectin and is responsible for the hemagglutinating activity expressed on formaldehyde-fixed thrombin-treated platelets. The direct effect of highly purified TSP on thrombin-induced platelet aggregation was studied. It was observed that aggregation of gel-filtered platelets induced by low concentrations of thrombin (less than or equal to 0.05 U/ml) was progressively inhibited by increasing concentrations of exogenous TSP (greater than or equal to 60 micrograms/ml). However, inhibition of platelet aggregation by TSP was not observed when higher than 0.1 U/ml thrombin was used to activate platelets. To exclude the possibility that TSP inhibits platelet aggregation by affecting thrombin activation of platelets, three different approaches were utilized. First, by using a chromogenic substrate assay it was shown that TSP does not inhibit the proteolytic activity of thrombin. Second, thromboxane B2 synthesis by thrombin-stimulated platelets was not affected by exogenous TSP. Finally, electron microscopy of thrombin-induced platelet aggregates showed that platelets were activated by thrombin regardless of the presence or absence of exogenous TSP. The results indicate that high concentrations of exogenous TSP (greater than or equal to 60 micrograms/ml) directly interfere with interplatelet recognition among thrombin-activated platelets. This inhibitory effect of TSP can be neutralized by anti-TSP Fab. In addition, anti-TSP Fab directly inhibits platelet aggregation induced by a low (0.02 U/ml) but not by a high (0.1 U/ml) concentration of thrombin.(ABSTRACT TRUNCATED AT 250 WORDS)

Plasma thrombospondin as an indicator of intravascular platelet activation in patients with vasculitis.

The use of plasma thrombospondin (TSP) concentration was investigated as an indicator of intravascular platelet activation. Patients (n = 20) with diseases that have known vasculitis were included in the study. The range and the mean of plasma TSP concentrations of patients with vasculitis were 117 ng/ml to 6500 ng/ml and 791 +/- 1412 ng/ml (mean +/- SD); the range and the mean of plasma TSP concentrations of control individuals (n = 33) were 13 ng/ml to 137 ng/ml and 59 +/- 29 ng/ml. When plasma TSP concentrations were correlated with plasma concentrations of another platelet activation marker, beta-thromboglobulin (beta-TG), it was found that the TSP concentration increased exponentially as the plasma beta-TG level rose. A positive correlation between plasma levels of plasma TSP and serum fibrin degradation products was also observed. The results suggest that platelets are the primary source of plasma TSP in patients with various vasculitis and that plasma TSP can be a better indicator than beta-TG to assess intravascular platelet activation due to its longer circulation half life.

Measurement of tissue factor activity in whole blood.

High circulating levels of the procoagulant molecule tissue factor (TF) are associated with thrombosis in a variety of diseases including unstable angina, cancer, and sepsis. Currently, there are no clinical assays to measure the level of TF activity in whole blood. We present an assay called Tissue Factor Clotting Time ("TiFaCT") that detects fibrin formation in human blood. The mean baseline clotting time in a healthy population was 472 +/- 94 s (mean +/- SD, n = 150). Bacterial lipopolysaccharide (LPS or endotoxin) shortened the clotting time in a time-dependent manner. Inhibitory anti-TF antibodies prolonged the clotting time of LPS-stimulated blood, indicating that the shortened clotting time was due to induction of TF expression. Patients with unstable angina had shortened mean baseline clotting time (284 +/- 86, n = 13) compared with healthy volunteers (474 +/- 98, n = 30), suggesting that these patients had elevated levels of circulating TF. The TiFaCT assay should prove clinically useful in quantifying the levels of circulating TF in patients at risk of thrombosis.

Mechanisms for genetically predetermined differential quantitative expression of HLA-A and -B antigens.

Previous studies showed that different HLA-A and -B antigens are differentially expressed in cells. Their relative quantities are genetically predetermined and inherited according to Mendelian laws. To investigate mechanisms responsible for this differential expression, a correlation study between the relative quantities of different HLA-A and -B proteins and their mRNA levels in eight different HLA-phenotyped lymphoblastoid cell lines (LCLs) were performed. The results show proportional correlation in all the studied cell lines except those that are positive for HLA-A24. Study of the turnover of HLA antigens reveals that different HLA-A and -B antigens are proportionally degraded. Measurement of the relative quantities of HLA-A and -B mRNAs in six LCLs before and after treatment with 5,6-dichloro-1-beta-D-ribofuranosylbenzimidazole (DRB), an inhibitor of RNA polymerase II, demonstrates that HLA-A and -B mRNAs are proportionally degraded except slight differences in two LCLs. Measurement of the relative quantities of different HLA-A and -B pre-mRNAs in nuclei shows that they are not proportional to the relative quantities of their respective mature mRNAs in cytoplasm in four of six LCLs. These results indicate that combinations of different regulatory steps which include gene transcription, pre-mRNA splicing and mRNA degradation are involved in the genetically predetermined quantitative differential expression of HLA-A and -B antigens. Transcription of HLA genes and splicing of HLA pre-mRNAs appear to be the dominant regulatory steps.

Genetic predetermination of quantitative expression of HLA antigens in platelets and mononuclear leukocytes.

In view of the potential functional importance of quantitative expression of HLA antigens, a series of studies were conducted to determine the relative quantities of specific HLA-A and -B antigens expressed in MNLs and platelets of HLA-phenotyped family members and unrelated individuals. An mAb that reacts with a well-defined monomorphic epitope in the alpha 3 domain of the heavy chains of HLA molecules was developed and used to quantify each HLA-A or -B antigen on western blots of IEF gels. The results of these studies demonstrated that the relative quantities of HLA-A and -B antigens in platelets and MNLs of an individual did not change over time. Further studies showed that the relative quantities of HLA-A and -B antigens for haplotypes shared among first-degree relatives were always the same and followed Mendelian inheritance. In contrast, the relative quantities of HLA-A and -B antigens for a haplotype shared by unrelated individuals varied significantly. All these findings support the hypothesis that the quantitative expression of HLA antigens is genetically predetermined and may play important roles in determining disease susceptibility and severity.

Proportional amplification of individual HLA-A and -B antigens during upregulated expression of total class I HLA molecules.

Our recent studies demonstrated that each specific HLA-A or -B antigen is not expressed in equal quantity in cells of an individual and that the relative amounts of different HLA-A and -B antigens are genetically predetermined following Mendelian laws. These findings suggest the potential genetic importance of varied quantitative HLA expression on target cells in determining the sensitivity to cytotoxic T lymphocytes. It would be important to know whether the amounts of different HLA antigens are differentially or proportionally amplified after upregulated expression of total HLA antigens. We have therefore determined the effects of IFN treatment, EBV transformation, and influenza virus infection on the quantitative expression of total HLA antigens and the relative quantities of different specific HLA-A and -B antigens in human fibroblasts cell line and peripheral blood mononuclear leukocytes. In contrast to earlier studies using the transfected HLA genes, our results show that different individual HLA-A and -B antigens are proportionally and not differentially amplified during upregulated expression of total class I HLA molecules. This finding indicates that the genetic predetermination of varied quantitative expression of HLA antigens may play a role in influencing antiviral immunity and disease susceptibility.

Induction of erythrocyte HLA expression during interferon treatment and HIV infection.

Although HLA antigens are present on the surface membrane of most cells, erythrocytes express little or no HLA. Occasionally red cells from normal individuals or patients with certain diseases express elevated levels of these molecules. The reasons for such variations are currently not understood. We report here that the expression of very high levels of HLA on erythrocytes occurs in response to interferon alpha given as a therapeutic agent for viral hepatitis. Increased expression became apparent after the second or third week of treatment, peaked at 3-4 months, and decreased at the end of the treatment period. This chronology suggests that elevated HLA expression is originated during erythropoiesis and persists throughout the lifetime of the erythrocyte. Furthermore, erythrocyte HLA expression did not correlate with changes of plasma HLA or beta 2-microglobulin concentrations and was not affected by in vitro chloroquine treatment, ruling out the possibility that HLA was adsorbed from plasma. Increased expression of HLA on erythrocytes was also demonstrated in patients infected with the human immunodeficiency virus, a disease in which increased production of endogenous interferon has been previously documented. We conclude that high HLA expression in red cells occurs in response to persistent interferon stimulation. Further studies will determine if this effect can also be produced by interferon tau or other factors.

Effect of HLA phenotype and gender on expression of various forms of class I HLA in plasma.

To understand the complexity of plasma HLA antigens, the distribution of different molecular weight forms of class I HLA in plasma was investigated in 44 HLA-phenotyped and unrelated individuals. Plasma class I HLA were immunoprecipitated by using the W6/32 anti-HLA monoclonal antibody, separated by SDS-polyacrylamide gel electrophoresis and characterized by immunoblotting with the HC-10 monoclonal antibody. Four different forms of HLA heavy chains (HLA-HC) with relative molecular masses of 44, 39, 36, and 34 kd were detected. Plasma samples from all individuals contained 44 and 36 kd HLA-HC, but varied as to the presence of 39 and 34 kd HLA-HC. Eighteen percent of the individuals did not have any detectable class I HLA with 39-kd heavy chains in their plasma and 61% did not have plasma class I HLA with 34-kd heavy chains. Thus, four different distribution patterns were identified for plasma class I HLA among all individuals included in our study. The distribution patterns in four different individuals were evaluated quarterly and remained unchanged during 1 year follow-up. A significant association of absence of 39-kd plasma class I HLA-HC with female gender (p less than 0.05) and HLA-B7 phenotype (p less than 0.00015) was also found. Further pedigree analyses of four families of HLA-B7-positive and 39-kd HLA-HC-negative probands indicated that genetic factor(s) other than those associated with HLA-B7 allele and female gender is involved in regulating the expression of the plasma class I HLA with 39-kd heavy chains.

Evaluation of individual specificities of class I HLA on platelets by a newly developed monoclonal antibody.

In order to quantify each specific HLA-A or -B antigen on platelets, a monoclonal antibody against HLA heavy chains was developed and designated as 2F2 monoclonal antibody. This monoclonal antibody reacted on Western blot with platelet HLA from each of 10 individuals with different HLA phenotypes and precipitated all 35S-methionine-labeled HLA-A and -B antigens from three different Epstein-Barr Virus--transformed lymphoblastoid cell lines. The results indicate that the 2F2 monoclonal antibody recognizes an epitope shared by different HLA-A and -B antigens. The quantitative variation of specific HLA antigens on platelets was then studied in nine different donors by isoelectric-focusing gel electrophoresis and immunoblot using the 2F2 monoclonal antibody. The results of our studies showed that the shared HLA antigens such as A2, B35, and B62, varied three- to fivefold among different individuals and individual HLA-A or -B antigen was not equally expressed on a person's platelets. The relative quantities of specific HLA-A and -B antigens on lymphocytes were also noted to be the same as those on platelets. The finding suggests that differential expression of HLA specificities may not be restricted to platelets but is a more general phenomenon including other nucleated cells.

Association between HLA phenotype and HLA concentration in plasma or platelets.

To understand the relationship between HLA phenotype and plasma or platelet HLA better, concentrations of plasma and platelet HLA were measured in 215 individuals of known HLA phenotypes. Precise quantitation of HLA antigens was achieved by means of an enzyme-linked immunoassay using the W6/32 monoclonal antibody and purified HLA molecules. The mean plasma and platelet HLA concentrations were 2.04 +/- 1.67 micrograms/ml (+/- SD, n = 215) and 11.28 +/- 4.65 fg/cell (+/- SD, n = 213), respectively. Statistical analysis of associations between HLA phenotypes and plasma HLA revealed that the mean plasma HLA concentration of individuals with HLA-A23 or HLA-A24 was 1.4 (p less than 0.002) or 1.9 (p less than 0.001) times higher than those without these two HLA antigens. Furthermore, the mean plasma HLA concentration of individuals who have HLA-A26 was 25% less than those without HLA-A26 (p less than 0.05). In contrast, the only association between HLA phenotypes and HLA concentrations of platelets was observed in HLA-B7-positive individuals. The mean platelet HLA concentration of HLA-B7 individuals was 27% higher than those without HLA-B7 (p less than 0.005). This finding is in accordance with previous observations made on red blood cells. The results indicate that the HLA concentrations in plasma are regulated, at least in part, by genetic factors that are different from those regulating platelet HLA.