Arch Dis Child: first published as 10.1136/adc.55.2.130 on 1 February 1980. Downloaded from

Archives of Disease in Childhood, 1980, 55, 130-133

Proteus vulgaris agglutination by cystic fibrosis sera

A WANG, M W TURNER, J F SOOTHILL, J 0 WARNER, AND A P NORMAN Department ofImmunology, Institute ofChildHealth, andRespiratory Unit, Hospitalfor Sick Children, London summARY The factor in sera ofpatients with cystic fibrosis (CF) and their parents which agglutinates vulgaris has characteristics similar to those of IgG antibody to this organism. Sera of patients without CF who have P. vulgaris agglutinate the organism similarly. At present there are too many false-positives in a control population for the test to be widely useful for heterozygote identification.

Sera of patients with cystic fibrosis (CF) and their drop of bacterial suspension and one drop of the parents contain a factor which inhibits cilial motility serum were dispensed by Pasteur pipette on to a in a variety of assay systems including rabbit microscope slide, covered with a coverslip, and tracheal explants,1 oyster,2 or water mussel gills.3 examined for agglutination at x 400 magnification Cohen and Daniel4 reported that suspensions of at 2-minute intervals for 25 minutes when the result the flagellate bacterium Proteus vulgaris were was recorded as large agglutinates (>80 Vpm), small agglutinated by such sera. agglutinates (<80 Vm), or no agglutination. We also We have investigated the mechanism and specificity used suspensions of , Escherichia of the P. vulgaris agglutinating assay and its applica- coli, , and Klebsiella tion as a screening test for CF gene heterozygotes in pneuimoniae. genetic counselling. The effects of a variety of treatments on the agglutination of CF sera were studied: these were Materials and methods repeated freezing and thawing, heating at 560C and 700C for one hour, heparin, incubation in acid Sera were obtained from 50 CF patients, from 57 of medium (pH 4-5 or 2 35), and precipitation of their parents (obligate heterozygotes), from 51 immunoglobulins with anti-immunoglobulin anti- healthy adults (medical and other staff of the sera (anti-IgM and anti-IgG, Wellcome Labora- http://adc.bmj.com/ hospital), from 6 children infected by proteus tories). The agglutinated by patients' sera (urinary, respiratory, or septicaemic infections), from were studied by indirect immunofluorescence, using 3 of their parents, from 12 children with asthma, and fluorescence-labelled sheep antihuman IgM and from 38 children in hospital for various reasons. We IgG (Wellcome Laboratories) at 1/10 dilution. also studied unstimulated whole mouth saliva of 23 Patients' sera were separated on a Sephadex G200 CF patients and 25 of their parents, and nasal column equilibrated with Tris-HC1 buffer pH 7-6. secretions from 2, urine from 8, and small intestinal IgG was prepared from patients' sera by DEAE juice from 11. The specimens were either tested cellulose batch chromatography6 using Whatman on October 2, 2021 by guest. Protected copyright. immediately after collection, or stored at -200C. DE-52 cellulose. The P. valgaris strain used was cultured from a Cilial motility tests were done with the gills of the child in the hospital, and was maintained at room fresh water mussel dreissensia.3 Gill segments were temperature on nutrient agar. Before assay it was cut with scissors from freshly collected animals, and cultured in 20 ml nutrient broth at 370C for 5 hours placed in closed moist plastic dishes containing when an actively motile log phase was attained. One 0 05 ml plasma and 0 10 ml distilled water; motility Department ofImmunology, Institute of Child Health was observed intermittently for 6 hours using phase M W TURNER, senior lecturer contrast illumination ( x 300). Inhibition of mobility J F SOOTHILL, professor ofimmunology at 3 hours was classified as + +, inhibition between Respiratory Unit, Hospital for Sick Children 3 and 6 hours as +, and no inhibition at 6 hours as A P NORMAN, consultant paediatrician Brompton Hospital, London negative. J 0 WARNER, consultant paediatrician H8pital Tenon, Paris, France Results A WANG, research fellow Most CF sera agglutinated P. vulgaris suspensions 130 Arch Dis Child: first published as 10.1136/adc.55.2.130 on 1 February 1980. Downloaded from

Proteus vulgaris agglutination by cystic fibrosis sera 131 within 3 or 4 minutes. Sera from obligate CF P. mirabilis, though less strongly than the P. vulgaris heterozygotes agglutinated them rather more slowly (Table 2). The agglutination of the other bacteria did and most sera from healthy adults had not not closely parallel that ofP. valgaris. agglutinated them by 25 minutes. The agglutinate These findings suggest that the factor is an formed by most CF sera were large (>80 V.m in agglutinating antibody to P. vulgaris, an organism diameter), those formed by the sera from the CF which sometimes infects CF patients, and to which parents were smaller-normally 10 to 40 F±m in their parents may well have excessive contact and diameter. In a blind study, using this classification, a perhaps special vulnerability. This view is supported significant difference between the three groups of by the observation that the sera of the youngest sera was obtained (Table 1). CF patients differed patient (aged 3 months) and those of both his parents significantly from controls whether analysed for failed to agglutinate this organism. production of large agglutinates or agglutinates of Sera from 5 out of 6 children with proteus any size (X2>30, P<0 0005 for both); obligate infection agglutinated P. vulgaris quickly into large heterozygotes differed from controls for agglutinates agglutinates, as did the serum from one mother; some ofany size (X2=19 * 3, P<0 °0005). of the children with asthma and miscellaneous 38 % of healthy adults produced agglutinates, diseases did so too (Table 3). Sephadex and DEAE including large ones in 14%. Only 5% of such separation was consistent with this being IgG and individuals would be expected to be heterozygous for IgM antibody. The urine of8 CF patients positive for the CF gene,6 so there are clearly mechanisms other serum agglutination did not agglutinate P. vulgaris than the CF gene leading to agglutination. either neat or after 50-fold concentration by ultra- As reported by Cohen and Daniel4 the agglutinates filtration. Saliva from 13 of 23 CF patients withstood repeated freezing, thawing, and heating agglutinated P. vulgaris, 6 of them with large (at 560C and 700C) for one hour. Treatment of the agglutinates, but only 6 of 18 of their parents, and serum with heparin did not prevent agglutination. only one of 10 healthy adults gave small agglutinates The agglutinates were partially destroyed at pH only (Table 4). In general, individuals gave both 2 - 35 but not at pH 4 v 5. saliva and serum activity, but the saliva of one CF The agglutinating factor in CF sera was not mother whose serum was negative did agglutinate P. precipitated by dialysis against distilled water and so vulgaris. Three out of the 11 small intestinal juice was shown to be a pseudoglobulin. After Sephadex samples from CF children agglutinated P. vulgaris. G200 gel filtration of the CF sera, agglutination was The relationship between the dreissensia gill cilial detected in the middle (IgG-rich) peak in all patients immobilisation test and P. vulgaris agglutination in but, in some, weak activity was also found in the first (IgM-rich) peak. IgG was prepared from the serum Table 2 Agglutination of5 bacterial species by sera http://adc.bmj.com/ ofa CF child by DEAE cellulose exchange chromato- from 3 CFpatients, 2 CF heterozygotes, and 2 healthy graphy; the activity remained with the IgG peak. subjects Immunofluorescence examination showed that the Proteus Proteus Escheri- Klebsiella Pseudo- agglutinates of P. vulgaris produced by CF sera vulgaris mirabilis chia coli pneu- monas contained IgG and occasionally IgM. Immuno- moniae aeruginosa precipitation with anti-IgG antibody greatly reduced CF patients the agglutination and anti-IgM partly reduced it, 1 ++ + _ _ _ 2 + + + - _ on October 2, 2021 by guest. Protected copyright. but anti-IgA had no effect. 3 + + + - _ Sera from 3 CF patients, 2 CF mothers, and 2 CF mothers 1 + - _ _ _ healthy adults were studied for agglutination of 5 2 + ++ - + other flagellate bacteria. The same sera agglutinated Healthy controls 1 - - - - + 2 + + + - + Table 1 Agglutination ofP. vulgaris by sera of CF patients, obligate heterozygotes, and healthy adult Table 3 P. vulgaris agglutination by patients with controls at 25 minutes diseases other than CF CFpatients Obligate Healthy adult Hospital Asthma Proteus Parent of (n = 50) heterozygotes cOhttols children (n = 18) proteus (n = 57) (n = 51) (n = 38) (n= 8) infected children Large agglutinates (n = 1) (>80 gm) 36 7 7 Small agglutinates Large agglutinates 6 3 7 1 (<80 pm) 11 38 12 Small agglutinates 17 8 0 0 No agglutination 3 12 32 No agglutinates 15 7 1 0 Arch Dis Child: first published as 10.1136/adc.55.2.130 on 1 February 1980. Downloaded from

132 Wang, Turner, Soothill, Warner, and Norman Table 4 Agglutination ofP. vulgaris by saliva of CF reacting organism. Bowman et al.7 similarly failed to homozygotes and heterozygotes separate the cilial immobilisation factor from IgG CF patients CF parents Healthy adults although they subsequently reported8 that CF (n = 23) (n = 23) (n = 10) fibroblast cultures released a protein which caused Large agglutinates 6 0 0 cilial immobilisation and suggested that this might Small agglutinates 7 6 1 associate with IgG in vivo. This concept requires No agglutinates 10 17 9 further study. Wilson et al.9-10 reported abnormal cationic protein bands after isoelectric focusing of Table 5 Association between P. vulgaris sera from CF homozygotes and heterozygotes, but agglutination and cilial immobilisation test in sera of did not directly investigate whether these were IgG CFpatients and their parents or not; their findings are therefore not incompatible P. vulgaris Cilial immobilisation with ours. agglutination As CF patients are often infected with flagellate CF patients CF parents bacteria, including P. vulgaris, and it is likely that ++ + - ++ + - their parents are therefore overexposed to such a Large agglutinates 5 2 0 1 1 0 contact, the test may simply be detecting antibody Small agglutinates 0 1 1 2 1 1 to these organisms, occurring as a tertiary effect of No agglutinates 0 0 1 0 1 1 the disease in both the patients and their parents, as close contacts rather than as a direct effect of the CF patients and their parents is shown in Table 5. gene. It is also possible that the incidence we Analysis of positive or negative reactions for the two observed in our healthy population might be tests on children and parents combined shows a misleadingly high, since they, as hospital workers, highly significant relationship (P = 0 004, Fisher's may be abnormally exposed to such infections. exact test). The one child without CF who had Since bacterial flagellae and cilia of other organisms proteus infection and whose serum agglutinated P. are structurally similar, it is also possible that the vulgaris gave an intermediate value for cilial cilial immobilisation tests result from the same immobilisation. sensitisation, if there is antigenic cross-reaction between these structures. Discussion The close relationship between the P. vulgaris agglutinating test and the cilial immobilisation test, These results confirm the report that sera from and the negative results in the one CF infant studied patients with CF and their parents agglutinate and his parents supports this view, but the finding

P. vulgaris more frequently, quickly, and strongly that agglutination of P. vulgaris was closely related http://adc.bmj.com/ than do sera from healthy subjects.4 It is closely to that of P. mirabilis but not to agglutination of associated with the previously reported immobilisa- other flagellates is to some extent against it. More tion of mammalian or invertebrate cilia.1-3 The work on such possible cross-reactions is needed. observation that parents (obligate heterozygotes) Before these tests can have any value in identifying gave intermediate values raised the possibility that heterozygotes as family members at risk, for the factor was closely related to the primary gene screening their potential spouses, or for epidemio- product but there are other hypotheses for it. Our logical work (for example, the role of CF results suggest that the factor concerned is an IgG heterozygosity in atopy)1' we must know more on October 2, 2021 by guest. Protected copyright. antibody to P. vulgaris; the evidence included gel about their mechanism. If the abnormal incidence of filtration and ion exchange chromatography separa- P. vulgaris agglutination occurs as a result of tion, inhibition by anti-IgG antiserum, and immuno- abnormal contact with flagellate bacteria it would be fluorescence analyses of the agglutinates. Sera of useless; if it is an abnormal response to a normal patients with proteus infection similarly agglutinated contact with an environmental factor, it could have P. vulgaris and the factor concerned had similar some limited value as has been suggested for atopy,1' characteristics. Since few people become ill with although many false-positives would be expected. It proteus infection and 5% of the population are is only if it is closely related to the primary gene thought to be heterozygotes for CF,6 it is possible product that the tests could be ofreal use. that these individuals were heterozygotes and had an increased susceptibility to proteus, but the We thank the Thames Water Board for supply of frequency of agglutination in the healthy population the dreissensia, Dr A D Patrick, Mrs B A M Harvey, was too high for this to be the only explanation, and and Mrs E Feldmann for advice and help, and the it is possible that the test results from previous Leukaemia Research Board for financial support sensitisation to a proteus or to an antigenically cross- for A W. Arch Dis Child: first published as 10.1136/adc.55.2.130 on 1 February 1980. Downloaded from

Proteus vulgaris agglutination by cystic fibrosis sera 133 References 8 Bowman B H, Barrett D R, Matalon R, Danes B S, Bearn A G. Cystic fibrosis. Fractionation of fibroblast Spock A, Heick H M C, Cress H, Logan W S. Abnormal media demonstrating ciliary inhibition. Proc Natl Acad serum factor in patients with cystic fibrosis of the Sci USA 1973; 70: 548-51. pancreas. Pediatr Res 1967; 1: 173-7. 9 Wilson G B, Jahn T L, Fonseca J R. Demonstration of 2 Bowman B H, Lockhart L H, McCombs M L. Oyster serum protein differences in cystic fibrosis by isoelectric ciliary inhibition by cystic fibrosis factor. Science 1969; focusing in thin-layer polyacrylamide gels. Clin Chim Acta 164: 325-6. 1973; 49: 79-91. 3 Besley G T, Patrick A D, Norman A P. Inhibition of the 10 Wilson G B, Fudenberg H H, Jahn T L. Studies on cystic motility of gill cilia of Dreissensia by plasma of cystic fibrosis using isoelectric focusing. I. An assay for detection fibrosis patients and their parents. J Med Genet 1969; 6: of cystic fibrosis homozygotes and heterozygote carriers 278-80. from serum. Pediatr Res 1975; 9: 635-40. 4 Cohen F L, Daniel W L. Effects of cystic fibrosis sera on Warner J 0, Norman A P, Soothill J F. Cystic fibrosis Proteus vulgaris motility. J Med Genet 1974; 11: 253-6. heterozygosity in the pathogenesis of allergy. Lancet 5 Stanworth D R. A rapid method of preparing pure serum 1976; 1: 990-1. 7y-globulins. Nature 1960; 188: 156-7. 6 Goodman H 0, Reed S C. Heredity of fibrosis of the pancreas. Possible mutation rate of the gene. Am J Hum Correspondence to Professor J F Soothill, Depart- Genet 1952; 4: 59-71. ment of Immunology, Institute of Child Health, 30 Bowman B H, McCombs M L, Lockhart L H. Cystic Guilford Street, London WC1N 1EH. fibrosis: characterization of the inhibitor to ciliary action in oyster gills. Science 1970; 167: 871-3. Received 2 March 1979 http://adc.bmj.com/ on October 2, 2021 by guest. Protected copyright.