Relationship of CD146 expression to secretion of interleukin (IL)-17, IL-22 and interferon-γ by CD4(+) T cells in patients with inflammatory arthritis

Expression of the adhesion molecule, CD146/MCAM/MelCAM, on T cells has been associated with recent activation, memory subsets and T helper type 17 (Th17) effector function, and is elevated in inflammatory arthritis. Th17 cells have been implicated in the pathogenesis of rheumatoid arthritis (RA) and spondyloarthritides (SpA). Here, we compared the expression of CD146 on CD4(+) T cells between healthy donors (HD) and patients with RA and SpA [ankylosing spondylitis (AS) or psoriatic arthritis (PsA)] and examined correlations with surface markers and cytokine secretion. Peripheral blood mononuclear cells (PBMC) were obtained from patients and controls, and synovial fluid mononuclear cells (SFMC) from patients. Cytokine production [elicited by phorbol myristate acetate (PMA)/ionomycin] and surface phenotypes were evaluated by flow cytometry. CD146(+) CD4(+) and interleukin (IL)-17(+) CD4(+) T cell frequencies were increased in PBMC of PsA patients, compared with HD, and in SFMC compared with PBMC. CD146(+) CD4(+) T cells were enriched for secretion of IL-17 [alone or with IL-22 or interferon (IFN)-γ] and for some putative Th17-associated surface markers (CD161 and CCR6), but not others (CD26 and IL-23 receptor). CD4(+) T cells producing IL-22 or IFN-γ without IL-17 were also present in the CD146(+) subset, although their enrichment was less marked. Moreover, a majority of cells secreting these cytokines lacked CD146. Thus, CD146 is not a sensitive or specific marker of Th17 cells, but rather correlates with heterogeneous cytokine secretion by subsets of CD4(+) helper T cells.

Does HLA-B27 influence the monocyte inflammatory response to lipopolysaccharide?

OBJECTIVES

How human leucocyte antigen B27 (HLA-B27) contributes towards arthritis susceptibility is still unclear, but effects on the response to bacteria unrelated to the classical antigen presenting role of B27 have been suggested. This study investigated whether HLA-B27 modulates the innate response to lipopolysaccharide (LPS), a component shared between all Gram negative bacteria that can trigger reactive arthritis.

METHODS

Pools of U937 transfectants expressing either HLA-B27, HLA-A2 or the expression plasmid alone were differentiated with phorbol 12-myristate 13-acetate and stimulated with LPS. Supernatants were analysed for tumour necrosis factor-alpha (TNF-alpha) secretion and the gene expression profiles of unstimulated and LPS-stimulated cells were determined by microarray analysis. Changes in gene expression that are indicative of an unfolded protein response (UPR) were also analysed by quantitative polymerase chain reaction (PCR).

RESULTS

TNF-alpha secretion, a biological marker of the inflammatory response to LPS, was not significantly different between U937-B27 and U937-control. No differences in gene expression between unstimulated U937-B27 and U937-control lines were detected. Both U937-control and U937-B27 exhibited a stereotypic response to LPS. Only one gene, OAS2, was differentially expressed by these cell lines, and this was confirmed by quantitative PCR. Analysis of XBP-1 splicing suggested that the UPR is induced following the LPS stimulation, but this increase was seen in all transfectants.

CONCLUSIONS

The expression of B27 does not profoundly alter the gene expression following LPS stimulation. Therefore, additional signals, such as those provided by cytokines or intracellular infection, may be required to reveal any influence of B27 expression on the inflammatory response.

Alpha beta but not gamma delta T cell clones in synovial fluids of patients with reactive arthritis show active transcription of tumour necrosis factor alpha and interferon gamma

OBJECTIVE

To compare the cytokine expression profile of three CD8+, three CD4+, and three gammadelta+ T cell clones all derived from the synovial fluids of three patients with reactive arthritis (ReA).

METHODS

Complementary DNA based microarrays containing the specific sequence of 56 cytokine transcripts were used for screening. Selected genes were confirmed by reverse transcriptase-polymerase chain reaction assay.

RESULTS

Microarray showed that transcripts encoding for interferon gamma and tumour necrosis factor alpha were expressed by all CD8+ and CD4+ T cell clones. However, gammadelta+ T cells predominantly expressed transforming growth factor beta2 and granulocyte monocyte-colony stimulating factor.

CONCLUSION

T lymphocyte clones from the joint of patients with ReA exhibit differential cytokine expression profiles. CD8+ and CD4+ T cells demonstrate a Th1 mediated profile, whereas gammadelta+ T cells show a more heterogeneous and less proinflammatory Th3 driven pattern.

Effects of Chlamydia trachomatis infection on the expression of natural killer (NK) cell ligands and susceptibility to NK cell lysis

Natural killer (NK) cells are an important component of the immediate immune response to infections, including infection by intracellular bacteria. We have investigated recognition of Chlamydia trachomatis (CT) by NK cells and show that these cells are activated to produce interferon (IFN)-gamma when peripheral blood mononuclear cells (PBMC) are stimulated with CT organisms. Furthermore, infection of epithelial cell lines with CT renders them susceptible to lysis by human NK cells. Susceptibility was observed 18-24 h following infection and required protein synthesis by the infecting chlamydiae, but not by the host cell; heat or UV inactivated chlamydiae did not induce susceptibility to NK cell lysis. CT infection was also shown to decrease the expression of classical and non-classical major histocompatibility complex (MHC) molecules on infected cells, thus allowing recognition by NK cells when combined with an activating signal. A candidate activating signal is MICA/B, which was shown to be expressed constitutively on epithelial cells.

Characterization of the humoral immune response to Chlamydia outer membrane protein 2 in chlamydial infection

Detection of antibodies to an outer membrane protein 2 (OMP2) by enzyme-linked immunosorbent assay (ELISA) by using either the Chlamydia trachomatis- or the Chlamydia pneumoniae-specific protein was investigated. OMP2 is an immunodominant antigen giving rise to antibody responses in humans infected with different C. trachomatis serovars (A to C and D to K) or with C. pneumoniae, which could be detected by OMP2 ELISA. OMP2 ELISA is not species specific, but antibody titers were usually higher on the homologous protein. The sensitivity of this assay was high but varied according to the "gold standard" applied. Levels of antibody to C. pneumoniae OMP2 as detected by ELISA seem to return to background or near-background values within a shorter period of time compared to antibodies to C. pneumoniae detected by microimmunofluorescence (MIF), making it more likely that positive results in ELISA reflect recent infection. Thus, OMP2 ELISA has distinct advantages over MIF and commercially available ELISAs and might be a useful tool for the serodiagnosis of chlamydial infection.

Recognition of the 60 kilodalton cysteine-rich outer membrane protein OMP2 by CD4(+) T cells from humans infected with Chlamydia trachomatis

T cell-mediated immunity is important in the control of chlamydia infection but chlamydia-specific T cells are also implicated in the inflammation and tissue damage which characterize chlamydia associated diseases. To investigate target antigens of the T cell-mediated immune response to chlamydia infection, Chlamydia trachomatis-specific CD4+ T cell clones were isolated from a patient with chlamydia-induced reactive arthritis. T cell immunoblotting indicated that an antigen of approximately 60 kilodaltons molecular mass was recognized, and recombinant 60 kilodalton cysteine-rich outer membrane 2 (OMP2) proved to be stimulatory. By using deletion constructs and synthetic peptides an epitope presented by HLA-DRB1*0401 was defined and proved to contain the nonamer peptide within the OMP2 sequence predicted to have the greatest binding affinity for DRB1*0401 The sequence of the epitope is conserved in all C. trachomatis strains but not in C. pneumoniae. Investigation of patients with acute urethritis and additional patients with sexually acquired reactive arthritis showed that OMP2-reactive T cells were readily detectable in peripheral blood and synovial fluid. Thus OMP2 is a target antigen of the T cell-mediated immune response to CT infection.

The recognition of HLA-B27 by human CD4(+) T lymphocytes

HLA-B27 transgenic animal models suggest a role for CD4(+) T lymphocytes in the pathogenesis of the spondyloarthropathies, and murine studies have raised the possibility that unusual forms of B27 may be involved in disease. We demonstrate that CD4(+) T cells capable of recognizing B27 can be isolated from humans by coculture with the MHC class II-negative cell line T2 transfected with B27. These CD4(+) T cells recognize a panel of B27-transfected cell lines that are defective in Ag-processing pathways, but not the nontransfected parental cell lines, in a CD4-dependent fashion. Inhibition of responses by the MHC class I-specific mAb w6/32 and the B27 binding mAb ME1 implicates the recognition of a form of B27 recognized by both of these Abs. We suggest that B27-reactive CD4(+) T cells may be pathogenic in spondyloarthropathies, particularly if factors such as infection influence expression of abnormal forms of B27.

Identification of Chlamydia trachomatis antigens recognized by human CD4+ T lymphocytes by screening an expression library

Identification of the immunogenic proteins that induce Chlamydia trachomatis (CT)-specific T cell responses is crucial to the development of protective vaccines and understanding the mechanisms of chlamydia-induced pathology. To characterize the targets of the human T cell response we have used chlamydia-reactive human T cell clones as cellular probes to screen a CT genomic library expressed in Escherichia coli using peripheral blood mononuclear cells to present antigens. The library was screened with three chlamydia-reactive T cell clones of unknown specificity and three novel stimulatory chlamydia antigens were identified. These E. coli recombinants were shown to express the chlamydia proteins, enolase, pmpD and CT579. Enolase and pmpD proteins were purified and shown to induce the proliferation of synovial fluid mononuclear cells isolated from the knee joints of patients suffering from chlamydia-associated reactive arthritis. We suggest that these stimulatory antigens are common targets of the T cell response in this group of patients. A greater understanding of T cell-mediated immunity in uncomplicated CT infection, and in patients with CT-induced chronic inflammatory disease (trachoma, salpingitis, arthritis) may identify the principal immune responses associated with immunopathology.

T lymphocyte lines isolated from atheromatous plaque contain cells capable of responding to Chlamydia antigens

Chlamydia pneumoniae infection is associated with atherosclerosis and the organism has been identified in arterial lesions. To determine whether T lymphocyte-mediated immune responses to Chlamydia antigens within plaque could contribute to pathogenesis, we have derived T cell lines from atherosclerotic plaques of 32 patients. Culture with IL-2 alone proved insufficient for cellular activation and expansion, but additional stimulation with phytohaemagglutinin (PHA) or recall antigens allowed consistent establishment of T cell lines. Furthermore, in cultures of approx. 500 tissue fragments, Chlamydia organisms proved as effective as other recall antigens in producing outgrowth of arterial T cells (20-25% wells produced T cell lines). Testing the antigen responsiveness of T cell lines showed that those derived using Chlamydia organisms were more likely to respond to Chlamydia (5/29+) than those isolated using other stimuli (6/69+ for PHA; 5/57+ for PPD and tetanus toxoid (TT)). However, lines responsive to each of the recall antigens were observed. Using recombinant Chlamydia antigens, some Chlamydia-specific T cell lines were shown to respond to OMP2 and/or hsp60. Those recognizing Chlamydia hsp60 did not cross-react with human hsp60, but human hsp60-responsive lines were also observed. Thus, atherosclerotic plaque tissue contains a variety of memory T lymphocytes, and amongst these are cells capable of recognizing Chlamydia antigens. In a C. pneumoniae-infected plaque, such T cells may be activated by local antigen and could contribute to the inflammatory process in the arterial wall through CD40 ligand expression and cytokine secretion.

Immune responses to Chlamydia antigens in atherosclerosis

Aim of this study was to isolate T lymphocytes from atheromatous plaques and to determine they respond to Chlamydia antigens. Atheromatous plaques from carotid endarterectomy patients, were cultured in vitro with the T cell growth factor, IL-2. This rarely allowed outgrowth of T cell lines. However, when combined with a mitogenic or antigenic stimulus to T cells, T cell lines were obtained from most patients, and from approximately 30% of replicate plaque tissue fragments. Chlamydia organisms were as effective in allowing the establishment of T cell lines as other recall antigens. T cell lines were tested for their ability to recognize antigens presented by autologous macrophages. Some lines responded to Chlamydia organisms, and also to the recombinant Chlamydia proteins hsp60 and OMP2. However, other lines recognized recall antigens. These results indicate that the atheromatous plaque contains memory T lymphocytes, and amongst the antigens they recognize are Chlamydia proteins. Stimulation of T cells was required to allow outgrowth in vitro, suggesting that the T cells were not in an activated state in vivo. However, since Chlamydia pneumoniae is present in the atheromatous plaque, activation of Chlamydia-reactive T cells by local antigen is a potential pro-inflammatory mechanism which could contribute to the pathogenesis of atherosclerosis.

Tracking antigen-specific human T lymphocytes in rheumatoid arthritis by T cell receptor analysis

The aim of this study was to use TCR sequencing as a tool to address the frequency of antigen specific T cells in different T cell compartments from a rheumatoid arthritis patient. We have previously established a clear link between T cell recognition of a specific Mhsp60 epitope and the amino acid sequence in the CDR3 region of the TCRB chain. This information was used to determine the frequency of these characteristic sequences in unmanipulated synovial fluid (SF), peripheral blood (PB) and hyperplastic lymph node of the same patient by amplification and sequencing. TCRBV sequences identical to those seen in antigen-specific clones, and closely related sequences, were readily identified in SF, where they represented approximately 1% of all T cells, but were absent from PB or lymph node. The prevalence of putative Mhsp60 specific T cells within the SFMC is much greater than previously suggested by limiting dilution assays. Thus, amplification and sequencing may prove a superior technique for tracking the frequency of antigen-specific T cells in different tissues and in a longitudinal fashion.

T cell responses to heat-shock protein 60: differential responses by CD4+ T cell subsets according to their expression of CD45 isotypes

We demonstrate that human T lymphocytes proliferate in vitro to highly purified human heat-shock protein 60 (Hu.hsp60). The response to this self Ag was confined to the CD45RA+ RO- T cell subset, with minimal responses by adult CD45RA- RO+ T cells. Experiments using keyhole limpet hemocyanin as a prototypic novel Ag, or tetanus toxoid as a recall Ag, were consistent with the notion that CD45RA+ RO- and CD45RA- RO+ T cell subsets can be designated as naive and memory cells, respectively; thus, responses to Hu.hsp60 were confined to the putative naive subset. In contrast, both CD45RA+ RO- and CD45RA- RO+ T cell populations proliferated to bacterial hsp60 from Mycobacterium leprae, Escherichia coli, or Chlamydia trachomatis. However, only CD45RA- RO+ (memory) T cells responded to a mycobacterial hsp60-derived peptide previously defined as a major bacteria-specific epitope. Experiments with cord blood T cells, which are CD45RA+ RO- and can be considered truly naive, showed that the peptide could elicit responses from naive T cells in vitro; cord blood cells also responded to Hu.hsp60. Since bacterial hsp60 Ags contain both conserved and nonconserved epitopes, we speculate that in vivo challenge with bacterial hsp60 will activate T cells capable of seeing either type of epitope, but only those that see nonconserved epitopes maintain the CD45RA- RO+ memory phenotype. However, T cells recognizing conserved epitopes, while not apparently being recruited to the memory pool, may nevertheless play a role in immunoregulation, particularly in the context of inflammation, when expression of Hu.hsp60 is increased.

Structural aspects of the interaction between heterogeneic human papillomavirus type 1 E4-specific T cell receptors and the same peptide/HLA-DQ8 complex

TCR usage has been studied in a panel of Th cell clones specific for the same peptide epitope (P N S Q D R G R P R R S D), derived from the human papillomavirus type 1 (HPV1) E4 protein, and restricted through HLA-DQ8. After identifying the V, D, and J genes used by the TCRs and sequencing across the V(D)J junctions, five different alpha-chain sequences and five different beta-chain sequences, comprising six independent clones, were identified. A structural model of our E4 peptide/HLA-DQ8 complex predicted that the guanidinyl side chain on the arginine residue at position 6 of the peptide could exist in different orientations. An intramolecular interaction between this arginine and the glutamine residue at position four appeared to control this orientation. Interacting HPV1 E4-specific TCRs would therefore have to recognize the complex in different conformations, and molecular modeling of the TCRs suggested that this could be achieved by changing the dimensions of the central pocket formed where the CDR3 loops of the TCR alpha- and beta-chains converge. It is known that interactions between bound peptide and amino acid residues lining the peptide-binding cleft of HLA molecules are important for determining the conformation and orientation of the peptide/MHC complex. The suggestion here that intramolecular interactions between amino acids of close proximity on the bound peptide are also important adds a further level of complexity to the mechanism by which TCRs interact with Ag.

Human gamma delta T-cell recognition of Yersinia enterocolitica

We have studied the human gamma delta T-cell response to Yersinia enterocolitica, a facultative intracellular bacterium which causes gastroenteritis and, particularly in human leucocyte antigen (HLA)-B27+ individuals, reactive arthritis (ReA). A marked proliferation of that cytotoxic gamma delta T cells is seen when Yersinia-infected lymphoblastoid cell lines or fixed intact Yersinia are added to cultures of mononuclear cells derived from the synovial fluid of ReA patients or from the peripheral blood of healthy donors. In contrast, heat-inactivated Yersinia fail to stimulate the gamma delta T-cell response. The gamma delta T-cell lines generated killed both autologous and allogeneic infected cell lines. Interestingly, a T-cell line generated from synovial fluid mononuclear cells (SFMC) killed infected autologous cell lines and a cell line matched for HLA-B27 less well than infected allogeneic target cells. gamma delta T-cell clones isolated from this line were found to express V gamma 9V delta 2 T-cell receptor (TCR) and also killed infected mismatched cells more efficiently than autologous targets. Moreover, from experiments using major histocompatability complex (MHC)-deficient cell lines, it was apparent that target cell recognition was MHC independent. Our results suggest that gamma delta T cells can be involved in immunity to Yersinia enterocolitica and should be taken into account when considering immunopathological mechanisms leading to reactive arthritis.

Stimulation of peripheral blood lymphocytes with Campylobacter jejuni generates a gammadelta T cell response in patients with Guillain-Barrésyndrome

In three patients whose Guillain-Barré syndrome (GBS) was preceded by gastrointestinal infection due to Campylobacter jejuni, gammadelta T cells were generated from peripheral blood in response to in vitro stimulation with C. jejuni. In one of the patients, where a diagnostic sural nerve biopsy was performed, gammadelta T cells were also isolated following culture of the nerve tissue. Studies with healthy volunteers and C. jejuni gastroenteritis patients also showed preferential enrichment for gammadelta T cells in peripheral blood cells stimulated with C. jejuni, although the response was significantly lower than that seen in GBS patients. In two out of three GBS patients and all of the controls, gammadelta T cell receptor (TCR) gene usage was shown to be Vgamma9/Vdelta2+. In the GBS patient where nerve-infiltrating gammadelta T cells were isolated, these and C. jejuni-specific peripheral blood cells had similar TCR gene usage, predominantly consisting of Vgamma5/Vdelta1+ cells. Sequencing the Vdelta1 products from nerve and peripheral blood showed similarities in CDR3 length, but the single Vdelta1 sequence obtained from nerve was not identified in peripheral blood. These results suggest that the generation of gammadelta T cells is part of a normal immune response to C. jejuni, which, in patients with GBS, may contribute to the pathogenesis of their inflammatory neuropathy.

Effect of polymorphism of the HLA-DPA1 chain on presentation of antigenic peptides

Human T-cell clones that recognize a peptide from mycobacterial heat shock protein 60 in the context of HLA-DP were found to be sensitive to changes in the DPA1 chain of the restricting element, optimal responses being seen with the combination HLA-DPA1*0201 and HLA-DPB*0301. HLA-DP dimers containing HLA-DPA1*01 were only able to present antigenic peptides to T-cell clones when peptides were present throughout the period of coculture of T cells with antigen presenting cells. In contrast the optimal HLA-DP dimer could also stimulate T-cell clones maximally when incubated with peptides for 1 h and then thoroughly washed. This suggests that the DPA1 polymorphism influenced the strength of binding of antigenic peptides to the HLA-DP dimer. Modeling studies identified amino acid 31 of DPA1 as the polymorphic residue most likely to account for this effect. This is the first demonstration that the relatively limited polymorphism displayed by DPA1 has functional consequences.

Marked conservation of complementarity-determining region 3 of the beta-chain of TCRs recognizing a mycobacterial heat shock protein 60-derived peptide with strong sequence similarity to human heat shock protein 60

The variable gene usage and sequence of human TCRs specific for a particular MHC/peptide combination have been investigated. The peptide comprises amino acids 456-466 of the 65-kDa Mycobacterium tuberculosis heat shock protein (hsp60), and is recognized in the context of HLA-DP. TCRs from both synovial fluid and peripheral blood (PB)-derived T cell clones used only five different V beta genes, three of which are closely related (V beta 6.7a, V beta 6.7b, and V beta 21.3). Among TCRs using these three genes there was marked conservation of the beta-chain sequence, whereby complementarity-determining region 3 (CDR3) contained an amino acid motif (*R*G*, amino acids 96-100) in association with either J beta 1.4 or J beta 2.5. These conclusions were strengthened by analysis of peptide-stimulated T cell lines that revealed not only TCR beta-chain sequences identical with those seen in T cell clones, but also additional beta-chains with similar CDR3 region sequences and J gene usage. In contrast, T cell lines derived by using IL-2 or a control peptide revealed variable usage of V beta and J beta genes; V beta 6.7a/b sequences from these lines and from freshly isolated PB did not contain the CDR3 motif noted in TCRs from Ag-specific T cells. We suggest that the remarkably limited diversity of TCRs noted in this study is a consequence of the similarity between the mycobacterial hsp60 peptide and the equivalent peptide from human hsp60, and reflects the trimming of the TCR repertoire required to maintain self-tolerance.

Marked conservation of complementarity-determining region 3 of the beta-chain of TCRs recognizing a mycobacterial heat shock protein 60-derived peptide with strong sequence similarity to human heat shock protein 60

The variable gene usage and sequence of human TCRs specific for a particular MHC/peptide combination have been investigated. The peptide comprises amino acids 456-466 of the 65-kDa Mycobacterium tuberculosis heat shock protein (hsp60), and is recognized in the context of HLA-DP. TCRs from both synovial fluid and peripheral blood (PB)-derived T cell clones used only five different V beta genes, three of which are closely related (V beta 6.7a, V beta 6.7b, and V beta 21.3). Among TCRs using these three genes there was marked conservation of the beta-chain sequence, whereby complementarity-determining region 3 (CDR3) contained an amino acid motif (*R*G*, amino acids 96-100) in association with either J beta 1.4 or J beta 2.5. These conclusions were strengthened by analysis of peptide-stimulated T cell lines that revealed not only TCR beta-chain sequences identical with those seen in T cell clones, but also additional beta-chains with similar CDR3 region sequences and J gene usage. In contrast, T cell lines derived by using IL-2 or a control peptide revealed variable usage of V beta and J beta genes; V beta 6.7a/b sequences from these lines and from freshly isolated PB did not contain the CDR3 motif noted in TCRs from Ag-specific T cells. We suggest that the remarkably limited diversity of TCRs noted in this study is a consequence of the similarity between the mycobacterial hsp60 peptide and the equivalent peptide from human hsp60, and reflects the trimming of the TCR repertoire required to maintain self-tolerance.

Identification of the epitope recognized by the human V beta 5-specific monoclonal antibody 42/1C1. Potential implications for disease therapy

A panel of T-cell clones was generated that was specific for amino acid residues 4-13 of the mycobacterial 65-kd stress protein. All the clones were found to express a member of the V beta 5 family, as defined by PCR. However, the clones could be differentiated on the basis of different staining characteristics with the mAb 42/1C1. This antibody is known to recognize both V beta 5.2 and V beta 5.3, as was the PCR primer pair used in the analysis. Sequencing of the TCRs revealed that those clones which were not stained by 42/1C1 expressed a previously unidentified member of the V beta 5 family. By comparing the sequences of the V beta 5 family members that are recognized by 42/1C1 with those that are not, we were able to identify a probable epitope for the antibody. It is also clear from our data that the TCRs of T cells recognizing identical MHC-peptide combinations, although very similar, may be differentiated by mAbs, thereby posing potential problems in any proposed disease therapy involving treatment with monoclonals.

cDNA cloning of porcine interleukin 2 by polymerase chain reaction

Porcine interleukin 2 (IL-2) cDNA was cloned by polymerase chain reaction (PCR), using primers derived from the corresponding bovine sequence. The resulting porcine DNA sequence encodes a 154 residue IL-2 primary translation product. Comparison of the mature, secreted form of porcine IL-2 with those of other species was carried out in an attempt to identify differences that might contribute to the observed differing species specificities.

Changes in serum lipoproteins and high-density lipoprotein composition during isotretinoin therapy

Oral isotretinoin has been reported to increase serum total triglycerides (TG), cholesterol (TC), phospholipids (TPL), apoprotein B (apo B), and to reduce high-density lipoprotein cholesterol (HDL-C). To investigate the effects of isotretinoin on HDL, we measured HDL-C, HDL phospholipids (HDL-PL), apoprotein A1 (apo A-1), and HDL-C subfractions (HDL2-C and HDL3-C) in 24 healthy, male patients receiving a 16-week course of isotretinoin (1.0 mg/kg/day) for treatment of severe acne vulgaris. Patients were placed on a constant diet and fasting lipid parameters were measured every 4 weeks. Analysis of the data from the 20 patients who completed the study confirmed the reported increase in TG, TC, LDL-C, apo B, and LDL-C/HDL-C (all p less than 0.01) observed during isotretinoin therapy. Reduction occurred in HDL-C (p less than 0.05) and HDL2-C (p less than 0.01) while HDL3-C remained unchanged, indicating that the effect of isotretinoin is on the HDL2-C subfraction. Apo A-1 and HDL-PL did not change significantly, suggesting that the reduction in HDL-C represents cholesterol depletion of the HDL particle rather than a reduction in HDL mass. After discontinuing isotretinoin, serum lipid parameters returned to baseline levels.