Prevalence of Yersinia plasmid-encoded outer protein (Yop) class-specific antibodies in patients with Hashimoto's thyroiditis

Objective   To investigate the prevalence of class-specific antibodies (IgG, IgA) to Yersinia enterocolitica plasmid-encoded outer proteins (Yops) in patients with diagnosed Hashimoto's thyroiditis.
Methods   Seventy-one patients with Hashimoto's disease, 464 healthy blood donors and 250 patients with non-postinfectious rheumatic disorders (matched controls) were tested for class-specific antibodies to Yops. Anti-Yop antibodies were determined by ELISA and Western blot.
Results   The prevalence of class-specific antibodies to Yops as determined by ELISA was 14-fold higher (20 of 71; 28.2%) in people with Hashimoto's thyroiditis than in the two control groups. These results were confirmed by the Western blot, with 16 positive sera, three equivocal and one negative.
Conclusions   There is strong clinical and seroepidemiologic evidence for an immunopathologic causative relationship between Yersinia enterocolitica infection and Hashimoto's thyroiditis. Further investigation concerning the mechanisms involved and the possible effects of antibacterial chemotherapy on the outcome of Hashimoto's disease is warranted.     

Prevalence of yersinia plasmid-encoded outer protein (Yop) class-specific antibodies in multitransfused Greek patients with thalassemic syndromes

Objective: To evaluate the prevalence of class-specific antibodies (G, A, M) to Yersinia enterocolitica plasmid-encoded outer proteins (Yops), in a closely followed multitransfused population of patients with thalassemia.
Methods: Sera from 408 β-thalassemic patients and 386 healthy blood donors used as controls were analyzed with the enzyme-linked immunosorbent assay (ELISA) for IgG, IgA and IgM antibodies to yersinia outer proteins. The Yop antigen for the ELISA was prepared using a plasmid-bearing wild-type strain of Y. enterocolitica of serotype 0:8.
Results: Anti-Yop IgG antibodies were detected in 84 out of 408 β-thalassemic patients (20.6%) compared with only eight out of 386 (2.1%) healthy blood donors. None of the sera of either group was positive for anti-Yop IgA or IgM antibodies. On evaluating patients with registered clinical and laboratory signs of a previous yersinia infection in the period from 1978 to 1996, we found that those with a positive aggiutination test for Y. enterocolitica infection at the time of manifestation showed a higher rate of persisting IgG seropositivity to Yops than those with positive culture and clinical signs only. A significant percentage (9.49%) of the seropositive patients had no registered data of a past Y. enterocolitica infection. There was remarkable persistence of anti-Yop IgG antibodies in the thalassemic population, even in patients infected during the early years of our study period (1978–80).
Conclusions: The results suggest that the determination of class-specific antibodies to Yops, which are specific antigens for the pathogenic yersiniae ( Y. enterocolitica, Y. pseudotuberculosis and Y. pestis ), in addition to its usefulness in the diagnosis of infection, will be a very sensitive and specific index for epidemiologic studies.     

Dendritic cell function is perturbed by Yersinia enterocolitica infection in vitro

Infection with Yersinia enterocolitica is the cause of intestinal or extraintestinal diseases. We investigated the role of dendritic cells (DC), the most potent antigen-presenting cell (APC), in the course of infection with Y. enterocolitica in vitro. For these studies, DC were isolated from human peripheral blood and infected with green fluorescent protein (GFP)-labelled Y. enterocolitica. Bacteria were found within DC by FACS analysis and viable bacteria could be cultured from lysed cells. Within 24 h after infection, DC upregulated CD83 and CD86 followed at day 3, indicating maturation of DC. In contrast, for MHC class II, a marked but transient downregulation was observed at day 3 after infection, and downregulation to a lesser extent for CD80 at day 5. To assess the immunostimulatory capacity of DC, viable infected and uninfected DC were incubated with autologous T cells in the presence of phytohemagglutinin A (PHA). T cell proliferation was significantly reduced at days 4-6 after infection but not thereafter, whereas nonpathogenic Escherichia coli was not able to mimick this suppressive effect of Y. enterocolitica. The same suppression could be observed when infected DC were used in a mixed leucocyte reaction with allogeneic T cells. Thus Y. enterocolitica is able to invade DC, does not induce necrosis or apoptosis, but affects maturation of DC. However, MHC class II-molecules are downregulated initially, which coincides with a diminished immunostimulatory capacity of DC infected with Y. enterocolitica. The diminished immunostimulatory capacity of DC following infection with Y. enterocolitica in vitro might impair or delay elimination of bacteria thereby contributing to pathogenesis of bacterial enteritis or extraintestinal manifestations such as reactive arthritis.     

Cross-reactive idiotypes in sera from patients with leprosy, lupus and Lyme disease and from healthy individuals.

Monoclonal IgM autoantibodies have previously been generated from a patient with lepromatous leprosy. Polyclonal anti-idiotypes raised against two of these monoclonal antibodies (8E7 and TH9) were used in an immunoassay to detect the presence of idiotype in human serum. The anti-idiotypes recognize different but overlapping sets of idiotypic determinants, some of which are present on antibodies which bind to Mycobacterium leprae. Sera were tested from 16 individuals with leprosy, 45 with systemic lupus erythematosus, 20 with Lyme disease, and 80 healthy subjects. Positive sera were detected in all groups (seven, two, three, and four, respectively). In most cases the serum bound to both anti-idiotypes, the idiotype being present in the IgM and/or IgG fraction. Levels of the two idiotypes varied independently of total serum IgG concentration and, in serial samples from one patient, independently of each other. The results indicate that 8E7 and TH9 may be representative of serum antibodies which are commonly expressed in leprosy, but may also be expressed in other diseases and in health; and they suggest that such serum antibodies are encoded by a widely shared set of variable region genes.     

No causal relationship between Yersinia enterocolitica infection and autoimmune thyroid disease: evidence from a prospective study

The objective of this study was to evaluate prospectively the relationship between Yersinia enterocolitica (YE) infection and the development of overt autoimmune hypo‐ or hyperthyroidism (study A) and the de novo occurrence of thyroid antibodies (study B). This was a prospective cohort study of 790 euthyroid women who were first‐ or second‐degree relatives of autoimmune thyroid disease (AITD) patients. Follow‐up was 5 years, with annual assessments. Study A was a nested case–control study in which YE serological status was measured between cases {subjects who developed overt hypothyroidism [thyroid stimulating hormone (TSH) > 5·7 mU/l and free T4 (FT4) < 9·3 pmol/l] or overt hyperthyroidism (TSH < 0·4 mU/l and FT4 > 20·1 pmol/l)} and matched controls. For study B, 388 euthyroid women without thyroid antibodies at baseline were enrolled. The YE serological status was compared between subjects who developed thyroid peroxidase (TPO)‐antibodies and/or thyroglobulin (Tg)‐antibodies at 4‐year follow‐up and those who remained negative. For study A, the proportion of subjects positive for Yersinia enterocolitica outer membrane protein (YOP) immunoglobulin (Ig)G or YOP IgA did not differ between cases and controls at baseline. One year before the development of overt hypo‐ or hyperthyroidism, the proportion of subjects with YOP IgG was not different between cases and controls, but YOP IgA were less prevalent in cases. For study B, de novo occurrence of TPO (or TPO‐antibodies and/or Tg‐antibodies) did not differ between subjects in whom YOP IgG were positive or negative at baseline. Neither persistence nor emergence of YOP IgG at 4‐year follow‐up was associated with the occurrence of TPO‐antibodies or Tg‐antibodies. Similar results were observed with respect to YOP IgA. YE infection does not contribute to an increased risk of thyroid autoimmunity.     

C-terminal region of outer surface protein C binds borreliacidal antibodies in sera from patients with Lyme disease

Borreliacidal antibodies specific for outer surface protein C (OspC) are induced shortly after infection with Borrelia burgdorferi. In this study, we identified the region of OspC recognized by immunoglobulin M (IgM) and IgG borreliacidal antibodies. Sera from patients with early Lyme disease were screened for borreliacidal activity specific for B. burgdorferi 50772 and OspC antibodies. Seven sera that contained similarly high titers of each response were then chosen randomly and adsorbed with OspC or a truncated OspC (OspC-Dra) containing the 50 amino acids nearest the carboxy terminus. Adsorption with OspC or OspC-Dra completely eliminated the borreliacidal activity in six (86%) of seven sera and significantly decreased the activity in the remaining serum (titer of 10,240 to 1,280). Moreover, OspC antibodies were no longer detected by OspC enzyme-linked immunosorbent assay or in a Western blot that contained native OspC. The findings confirmed that sera from patients with early Lyme disease contain high concentrations of IgM or IgG borreliacidal antibodies that bind a conserved region of OspC.     

Increased prevalence of antibodies to enteropathogenic Yersinia enterocolitica virulence proteins in relatives of patients with autoimmune thyroid disease

Infections have been implicated in the pathogenesis of a number of autoimmune diseases, and Yersinia enterocolitica (YE) might play a role in the development of autoimmune thyroid disease (AITD). Clinical evidence in support of this hypothesis has been inconclusive. We reasoned that looking earlier in the natural course of AITD might enhance chances of finding evidence for YE infection. Consequently, we determined seroreactivity against YE in subjects at risk of developing AITD, i.e. in 803 female relatives of AITD patients in self-proclaimed good health. As a comparison group we used 100 healthy women who participated in a program for reference values. IgG and IgA antibodies to virulence-associated outer membrane proteins (YOPs) of YE were measured by a specific assay. Serum thyroid peroxidase antibodies (TPO-Ab) as indicators of AITD were considered to be positive at levels of> 100 kU/l. The prevalence of YOP IgG-Ab was higher in AITD relatives than in controls (40.1% vs. 24%, P = 0.002), and the same was true for YOP IgA-Ab (22% vs. 13%, P < 0.05). Of the 803 AITD relatives, 44 had an increased or decreased plasma TSH, and 759 were euthyroid as evident from a normal TSH; the prevalence of YOP-Ab did not differ between these three subgroups. TPO-Ab were present in 10% of controls and in 27% of the AITD relatives (P < 0.001). The prevalence of TPO-Ab in the euthyroid AITD relatives was not different between YOP IgG-Ab positive and negative subjects (23.3% vs. 24.7%, NS), nor between YOP IgA-Ab positive and negative subjects (21.2% vs. 24.9%, NS). In conclusion, healthy female relatives of AITD patients have an increased prevalence of YOP antibodies, which, however, is not related to the higher prevalence of TPO antibodies in these subjects. The findings suggest a higher rate of persistent YE infection in AITD relatives. Susceptibility genes for AITD may also confer a risk for YE infection.     

Antibodies to the peptide from the plasmid-coded Yersinia outer membrane protein (YOP1) in patients with ankylosing spondylitis

Seventy-five Norwegian patients with ankylosing spondylitis (AS) were studied for class-specific antibody response against synthetic peptide, P81, representing the sequence of plasmid-coded outer membrane protein of Yersinia (YOP1) containing four amino acid homology (TDRE) with HLA-B27 sequence. Ten (16.7%), five (8.3%) and seven (11.2%) of 60 male AS patients showed elevated anti-YOP1 P81 antibody of IgA, IgG, and IgM class, respectively, whereas for each isotype only one (4%) of 25 healthy male controls was positive. Differences were not observed between female patients and controls. In all isotypes, antibody-positive patients were more frequently found in patients with active disease. The anti-YOP1 P81 antibody levels of the patients were generally not correlated with the antibody levels against the peptide representing the hypervariable region of HLA-B27 (B27 peptide). However, in one patient the antibody was shown to react with both peptides by cross-inhibition analysis. Overall, it appears that any causal relationship between YOP1 and pathogenesis of AS is not strong. Immunogenicity and cross-reactivity of the YOP1 region encompassing the TDRE sequence particularly at the T cell level require further study.     

Cross-reaction of anti-DNA autoantibodies with membrane proteins of human glomerular mesangial cells in sera from patients with lupus nephritis

Anti-DNA autoantibodies were thought to play a major role in the pathogenesis of lupus nephritis (LN). A recent study revealed that affinity-purified anti-DNA antibodies had a cross-reaction with human glomerular mesangial cells (HMC). However, whether the cross-reaction was antigen-antibody-mediated was unclear. The aim of the current study was to investigate the binding of anti-DNA antibodies to HMC membrane proteins and to characterize the target antigens. Affinity-purified IgG anti-DNA antibodies were purified by DNA-cellulose chromatography in sera from nine patients with biopsy-proven active lupus nephritis. In vitro cultured primary HMCs were disrupted by sonication and HMC membranes were obtained by differential centrifugation. The membranes of human umbilical vein endothelial cells (HUVEC), human proximal renal tubular epithelial cell line (HK2) and peripheral mononuclear cells (PMC) were obtained as controls. Binding of anti-DNA antibodies to the membrane proteins was investigated by Western blot analysis using soluble membrane proteins as antigens. Both HMC membrane and affinity-purified anti-DNA antibodies were treated with DNase I to exclude DNA bridging. All nine affinity-purified anti-DNA antibodies could blot the HMC membrane proteins, and there were at least three bands at 74 kDa, 63 kDa and 42 kDa that could be blotted. Among the nine IgG preparations, all nine (100%) could blot the 74 kDa band; eight (88.9%) could recognize 63 kDa and 42 kDa protein bands separately. After DNase treatment, the same bands could still be blotted by most affinity-purified anti-DNA antibodies. Affinity-purified anti-DNA antibodies could also blot similar bands on membrane proteins of other cells, but some bands were different. In conclusion, anti-DNA autoantibodies could cross-react directly with cell membrane proteins of human glomerular mesangial cells and might play an important role in the pathogenetic mechanism in lupus nephritis.     

Effect of cytokines on invasion and survival of Yersinia in primary human fibroblasts

Bacteria or bacterial antigen triggering reactive arthritis have been detected in inflamed joint tissue and fluid of patients. Although live yersiniae have not yet been found in joints of patients with Yersinia arthritis, early dissemination and propagation has been proposed. In this study, we investigated the influence of the proinflammatory cytokines interleukin- (IL-) 1β, tumor necrosis factor- (TNF-) α, the Th1 lymphokine interferon- (IFN-) γ, and the Th2 lymphokine IL-4 on the intracellular survival of Yersinia enterocolitica O.3 in primary human fibroblast cell monolayers as a model for joint tissue. Bacterial titers in infected cells decreased significantly and in a dose-dependent manner following treatment with IL-1β, TNF-α, or IFN-γ. The bactericidal effects of IL-1β and TNF-α were synergistic. In contrast, IL-4 significantly supported bacterial survival. In addition, IL-4 antagonized in part the bactericidal effect of TNF-α and IFN-γ. Although IL-1β, TNF-α, and IFN-γ accelerated killing of intracellular yersiniae the ratio of cells containing bacterial antigen did not differ from that in untreated cells. The differential effects of the investigated cytokines on intracellular survival of yersiniae may be of relevance for the development of Yersinia arthritis: enhanced production of IL-4 by synovial tissue may prolong the survival of yersiniae and persistence of antigen, and thus potentiate immune complex formation and inflammation. In conclusion these results show (1) that fibroblasts can take up virulent yersiniae as non-professional phagocytes and (2) that cytokines, found in the joints of patients with Yersinia arthritis, are able to affect the intracellular survival of yersiniae differentially. Received: 19 September 1998     

Naturally occurring antibodies in human sera that react with the iron-regulated outer membrane proteins of Escherichia coli

Sera from normal healthy human adults and infants, as well as sera from mice, rabbits, and guinea pigs, were examined by immunoblotting for naturally occurring antibodies reacting with outer membrane proteins of two Escherichia coli strains, O111 and O18. Some individuals had antibodies reacting very strongly with the iron-regulated outer membrane proteins, including the ferric-enterochelin receptor protein (Mr, 81,000), as well as with ompA. However, sera from infants contained predominantly antibodies to ompA; antibodies recognizing the iron-regulated outer membrane proteins were either absent or barely detectable. In human serum the antibodies were mainly of the immunoglobulin G class. No serotype-specific antibodies to the lipopolysaccharide of E. coli O111 or O18 were found in the sera tested.     

Kinetics of B cell response during Yersinia enterocolitica infection in resistant and susceptible strains of mice

In this study we analyze the B-cell response in murine yersiniosis. To this end, we determined whether polyclonal activation of B-lymphocytes occurs during infection of susceptible (BALB/c) and resistant (C57BL/6) mice with Y. enterocolitica O:8 and compared the immunoglobulin (Ig) isotypes produced in response to the infection by the two strains. The number of splenic cells secreting nonspecific and specific immunoglobulins was determined by ELISPOT. The presence of anti-Yersinia antibodies in serum was detected by ELISA. In both strains, the number of specific Ig-secreting cells was relatively low. Polyclonal B-cell activation was observed in both strains of mice, and the greatest activation was observed in the BALB/c mice, mainly for IgG₁- and IgG₃- secreting cells. The C57BL/6 mice showed a predominance of IgG_2a-secreting cells. The peak production of anti-Yersinia IgG antibodies in the sera of BALB/c mice was seen on the 28th day after infection. The greatest increase in IgM occurred on the 14th day. A progressive increase of specific IgG antibodies was observed in C57BL/6 mice up to the 28th day after infection while IgM increased on the 21st day after infection. The production of specific IgA antibodies was not detected in either BALB/c or C57BL/6 mice. We conclude that polyclonal activation of B lymphocytes occurs in both the Yersinia-resistant and Yersinia-susceptible mice and that the more intense activation of B lymphocytes observed in the susceptible BALB/c mice does not enhance their resistance to Y. enterocolitica infection.     

Characterization of a Recombinant Yersinia enterocolitica Lipoprotein; Implications for its Role in Autoimmune Response against Thyrotropin Receptor

Autoimmune Graves' disease (GD), which is characterized by hyperthyroidism, is mediated by autoantibodies to the thyrotropin receptor (TSHR). Yersinia enterocolitica (Y.e.) has been shown to produce a lipoprotein (LP) that can cross-react with the TSHR and thus can act as a potential trigger of thyroid autoimmunity. In this study, to further characterize LP, we cloned the LP gene from Y. enterocolitica and expressed a recombinant LP. This recombinant LP was mitogenic for C3H/HeJ (LPS hyporesponsive) B cells and induced production and secretion of significant levels of IL-6 from splenocytes. A mouse antibody generated against the recombinant LP cross-reacted with TSHR as shown by western blot analysis. FACS analysis of splenocytes from mice immunized with LP revealed that LP could induce increased expression of B7.1 and B7.2. The immunomodulatory effects of LP including up-regulation of B7.1 and B7.2 coupled with its ability to induce antibodies that can cross-react with the TSHR showed several potential mechanisms by which it can cause breakdown of self-tolerance to TSHR.     

Heterogeneity of Borrelia burgdorferi sensu lato demonstrated by an ospA-type-specific PCR in synovial fluid from patients with Lyme arthritis

Borrelia burgdorferi sensu lato, the etiological agent of Lyme borreliosis, has been divided into three genospecies: B. burgdorferi sensu stricto (OspA-type 1), B. afzelii (OspA-type 2) and B. garinii (OspA-type 3–7). Whereas in Europe B. afzelii (OspA-type 2) is predominant among human skin isolates and B. garinii (OspA-type 3–7) among human CSF isolates, some previous serological studies suggested that Lyme arthritis is also associated with B. burgdorferi sensu stricto in Europe. In the present study we designed ospA type-specific PCRs and identified four different ospA types associated with Lyme arthritis. Our study group consisted of 20 patients with positive serology (ELISA and immunoblotting) and clinical criteria for Lyme arthritis. B. burgdorferi DNA was detected in 13 patients and in none of 10 control patients from synovial fluid. We identified ospA-type 1 (26.6%), ospA-type 2 (33.3%), ospA-type 4 (6.6%) and ospA-type 5 (33.3%). Our conclusion is that in Europe B. burgdorferi sensu lato strains causing Lyme arthritis are considerably heterogeneous and that there is no prevalence of certain genospecies or OspA-types among this strains. Received: 14 May 1998     

Autoantibodies against cyclophilin in systemic lupus erythematosus and Lyme disease.

Autoantibodies against cyclophilin, a cyclosporin A binding protein, were detected in sera of 29 of 46 (63%) patients with systemic lupus erythematosus and 14 of 40 (35%) Lyme disease patients. The antibodies are directed against the denatured form of both the major and minor isoform of cyclophilin and can be demonstrated in Western blots. Some first-degree relatives of lupus patients also express these antibodies. They are specific for cyclophilin and are not the consequence of hypergammaglobulinaemia. Four monoclonal IgM antibodies from a patient with lepromatous leprosy also bound to cyclophilin. The generation of these antibodies may be of special interest because they are against a protein involved in the control of the immune system not known to be directly associated with DNA or RNA.     

Evaluation of a western blot method for the detection of Yersinia antibodies: evidence of serological cross-reactivity between Yersinia outer membrane proteins and Borrelia burgdorferi

Yersinia enterocolitica and Yersinia pseudotuberculosis have been identified as causative organisms of reactive arthritis in humans. We evaluated a Western blot assay which uses Yersinia outer membrane proteins as antigens for the detection of Yersinia antibodies as a replacement for the complement fixation (CF) assay. Clinical agreement, sensitivity, and specificity were determined by testing 19 positive and 21 negative serum samples by the CF assay, Western blot assay, and enzyme-linked immunosorbent assay (ELISA). The CF assay and ELISA were compared to the Western blot assay, which was the reference method used in this study. Sera with antibodies that could potentially cross-react with Yersinia were also tested by the Western blot assay. The agreement, sensitivity, and specificity of the CF method were 61%, 26%, and 95%, respectively; and those for the ELISA were 89%, 95%, and 82%, respectively. The prevalences of Yersinia antibodies in 50 healthy donors were 6% for immunoglobulin G (IgG), 2% for IgA, and 2% for IgM. Sera positive for Bartonella henselae, Brucella, Chlamydia pneumoniae, and Rickettsia rickettsii antibodies showed cross-reactivity by the Western blot assay. The highest cross-reactivity was observed with Borrelia burgdorferi; 5 of 11 (45%) specimens were cross-reactive by the IgM-specific assay. Overall, the Western blot assay performs acceptably and is more sensitive than the CF assay, warranting replacement of the CF assay in the laboratory. Due to the evidence of cross-reactivity, particularly with B. burgdorferi, which can cause an oligoarthritis similar to reactive arthritis, the diagnosis of reactive arthritis should be based on clinical findings and complete serologic analysis of the potential causative infectious pathogens.     

Yersinia outer proteins E, H, P, and T differentially target the cytoskeleton and inhibit phagocytic capacity of dendritic cells

Through Yersinia outer proteins (Yops) Yersinia disrupt the actin cytoskeleton of epithelial cells and macrophages, and this leads to a decreased capability of these cells to internalize bacteria. We examined the effects of different Yops of Y. enterocolitica serotype O8 on the cytoskeleton and phagocytic capacity of murine dendritic cells (DCs). DCs were infected with several Yersinia mutant strains deficient in one Yop or translocating only a single Yop. Analyses of infected DCs by microscopy showed that YopE, YopH and YopT cooperate to rapidly damage the actin cytoskeleton of DCs. Furthermore, microscopic analyses and gentamicin killing assays revealed that the maximum reduction of bacterial uptake was achieved by Yersinia mutant strains translocating only a single Yop (YopE or YopH) indicating that these Yops enable Yersinia to inhibit the phagocytic function of DCs.     

Lyme borreliosis in Sweden--diagnostic performance of five commercial Borrelia serology kits using sera from well-defined patient groups

Five commercial Borrelia serology kits available in Sweden were evaluated and compared for their diagnostic performance in sera from clinically well-characterized patient groups. With the clinically defined groups as the gold standard, i.e. without knowledge of antibody status in serum and cerebrospinal fluid, the diagnostic performance of the kits was compared and important differences in diagnostic usefulness were found. The kits from Abbot and DAKO, that often predict clinically relevant Borrelia infection and do not detect antibodies in sera from patients without strong suspicion of Borrelia infection, were considered the most useful in the population studied. This kind of validation study is an important part of good laboratory practice and should be performed by laboratories serving patient populations with varying endemicity of Borrelia.     

Serum antibodies to outer membrane proteins of Escherichia coli in healthy persons and patients with bacteremia

Antibodies to Escherichia coli outer membrane proteins in sera from healthy persons and from patients bacteremic with various enteric or nonenteric bacteria were measured by an enzyme-linked immunosorbent assay (ELISA). Outer membranes were prepared from E. coli O55. Serum was absorbed with E. coli O55 lipopolysaccharide and diluted 1:100 for immunoglobulin A (IgA) or IgM and 1:1,000 for IgG antibodies. Paired serum specimens were obtained from the 56 patients included in the study (the first specimen on the day of positive blood culture and the second specimen 8 to 12 days later) and compared with sera from blood donors (n = 50) as controls. On an average, the patients bacteremic with enterobacteria (n = 40) showed increased levels of antibodies of all three immunoglobulin classes in the first serum specimens and significantly higher levels in the second specimens compared with the controls, although with considerable case-to-case variation. Increased levels of IgG antibodies showed the best combination of diagnostic specificity (100%) and sensitivity (53%) for bacteremia caused by enteric bacilli. Mostly, the antibody response was directed against the major E. coli O55 outer membrane proteins at 81,000, 38,500, 33,500, and 7,500 molecular weights as shown by Western blot (immunoblot) analysis. Some of the patients bacteremic with nonenteric bacteria showed increased levels of IgA antibodies, but not of IgG or IgM antibodies. Cross-reactivity of the nonenteric blood culture isolates with the E. coli outer membrane preparation was not demonstrated. The cross-reactivity of the E. coli O55 outer membrane proteins with those of enteric bacilli of other genera was examined by absorption experiments. Western blots with serum absorbed with live E. coli O55 provided evidence that the epitopes of the outer membrane protein at 7,500 molecular weight were available for antibody binding at the bacterial surface, and that at least some of the epitopes of the 38,500- and 33,500-molecular -weight proteins were accessible to antibodies. The results suggest that an ELISA for the measurement of antibodies against cross-reactive outer membrane proteins from enteric bacilli may be useful in the diagnosis of serious infections caused by members of the family Enterobacteriaceae, and that antibodies to the major outer membrane proteins may have an immunobiological function.