AJEBAK 56 {Pt. 6) 71.3-725 (1978)
THE SOURCE OE IgM AND IgG IN THE DOG INTESTINE by R. J. HEDDLE" AND D. ROWLEY (From the Department of Microbiology and Immunology, University of Adelaide, Adelaide 5001, South Australia.) (Accepted for publication October 10, 1978.) Summary. The source of IgM ant! IgGg., ,, in the dog intestine was investigated, using labflled dog iinmuriogloinilins and thionic Thiry-Vella loops of dog small intestine. Clearance.s of IgM and IgGg,, h from serum into intestinal secretions were similar, reflecting transfer of a greater ma.ss of IgGgy ^ than of IgM. Only 20% of the intestinal IgM derived from .serum, irrespective of the marker u.sed. Almiit one third of the total intestinal IgG. of all sub-classes and accrued from all sources, was seruni-ilerived IgG^,, i,. This indicates that serum is a major contributor to clog intestinal IgG^,, ,,.
INTRODUCTION Secretory IgA is quantitatively the dominant immunoglobulin in the lumen of the small intestine, but receut studies indicate that IgM and IgG antibodies or their fragments may play an important role in the intestine (Horsfall et al, 1978). The quantities of IgM and IgG transported from serum into the small intestine have been estimated in several studies. Wernet et al (1971) found that more (radio-labelled) rabbit IgG than rabbit IgM or secretory IgA was transported from serum into ligated loop.s of rabbit intestiue after cholera toxin stimnlation. Tlie proportions of each class of intestinal immunoglobulin transported from serum were not estimated. The major defect witli this study is that the permeability of the intestiual loops may have been increased (De and Chatterje, 1953). Recently, Cripps, Husband and Lascelles (1974) showed, in sheep, that the intestinal IgG2, most of the intestinal IgCi and an uncertain proportion of the intestinal IgM derived from serum. The contributions of the possible pathways by which immunoglobulins might reach the intestine are still not well understood, and this study was designed to examine the source of IgM and IgG in the dog small intestine using what was hoped would be a physiologically intact model. " Present address; Department of Medicine, Royal Adelaide Hospital, University of Adelaide, Adelaide 5001, South Australia.
7X4
R. J. HEDDLE AND D . ROWLEY MATERIALS AND M E T H O D S
Transport was detemiined using Thiry-Vella loops of dog small intestine which allowed easy acce.ss to small intestinal secretions and excluded other gastrointestinal secretions. Marker inmninoglobuiin was injected intra\enously. Transport was determined 2-3 days later by measuring the rate at which marker immunoglobulin entered the Thiry-Vella loop during irrigation. Three labelled dog immunoglobulins were used:— ^^^I-IgM. i''*iT-IgG.,.| ,, and IgM antibody capable of killing V. cholerae in the presence of complement. The latter was used to control whether the raitio-labelling procedures had changed the transport characteristics of the proteins a.s reported hy Tengstrom (1965) and Waldmann and Strober {1969). Recovery of "vibriocidar' IgM antibody indicated the presence of IgM antibody molecules with a high level of functional integrity. Dogs Adult mongrels, weighing 10-35 kg. were obtained from the Central Animal House of this Uni\ersil\\ Chronic Thiry-Vella loops were prepared in the middle third of the small liowel (Shields, Mulholland and Elmslie, 1966). The loops were irrigated every 2-3 days with isotonic "irrigating .solution" (description follows). Inimtitioglobulim—preparation and purity Dog IgG.j^i ,, and IgM were purified by published methods (Heddle and Rowley. 1975h). Double imuiunodittusion against rabbit anti-whole dog .serum and goat antisera monospecific for dog IgM (anti-n), IgG (anti-y) and (secretory) IgA (anti-a) (Heddie and Rowley. 1975a; Heddle. 1976) revealed no impurities. Immunoglobulins were labelled with carrier-free ' ^U (Amersham, Buckinghamshire, England) using the chloramine-T technicjue of McConahey and Dixon (1966). Radiolabelled immunoglobulins were separated from free radio-iodine by gel filtration (Sephadex G-25). Resulting specific activities were approx. 10 nCi/mg. Serum was obtained from a dog wbich bad been immunised with washed live V. cholerae 569B (Neoh and Rowley. 1970). "Vibriocidal" IgM antibody was obtained by gel filtration of this serum (Heddle, 1976). The same procedure demonstrated that the vibriocidal activity in the immune serum was associated with IgM antibody. Immunoglobulins—measurement and efiaracterisation in sera and intestinal washings Single radial immunodiffusion (SRW). Total IgM, IgG and IgA were measured by SRID (Mancini. Carbonara and Heremans. 1965) using the respective monospecific antisera and appropriate standards (Heddle, 1976). The cibriocidal assay was based on that described by Neoh and Rowley (1970). The vibriocidal acti\ity of intestinal washings was characterised by incubating (.'37°, lh) washings before assay with equal voluuies of (a) 0-2 M mercaptoethanol (freshly diluted in phosphate buffered saline (PBS). pH 7-4). (b) V. cholerae 569B lipopolysaceharide (Steele, Chaicumpa and Rowley. 1974) at 40-100 ng/ml in PBS. or (c) PBS. Samples were dialysed against PBS and titrated in duplicate using doubling dilutions. Coiitiiing of radioactivity was performed in a Packard Autogamma Spectrometer. Samples were counted to a niinimuni of 1,000 counts greater than background and counts were corrected for decay of radioacti\ity. Counting errors were estimated using standard formulae (Nuclear-Chicago Corporation. 1967). Immunopredpitation of radioactivity was achieved using monospecific goat anti-dog sera which had been prepared and linked covalently to agarose beads as described (Heddle and Rowley, 1975 a and b). Duplicate aliquots of samples were e.\posed (room temperature, 30 min) to either an excess of the appropriate anti-dog heavy chain immunoadsorbent or an
TRANSPORT OF DOG IMMUNOGLORULINS
715
equal quantity of a beterologoiis iniuumoaclsorbent. Followinjj; this, the inimunoadsorbents were wasbed live times in ice-cold PBS (pH 7-1) and counted. If "I" represents tbe counts which were present in the sample. "S" the counts on tbe washed homologous ininuuioadsorbent, and "C" the eounts on the washed heterologous imnuinoadsorbent, tbe proportion of the available eoimts adsorbed specificaliy was taken as S-C divided by I-C Binding to tbe heterologous immunoadsorbent was assumed to be representative of nonspecific binding to the homologous immunoadsorbeut. The effect of counting errors on the accuracy of this calculation was estimated using published formulae (Nuclear-Chicago Corporation. 1967; Arinitage, 1971). Trichloruacetic aciil preparation and gel filtration. Some radioactivf samples were analysed further by precipitation with an eqna! volume of cotd 2O'-6 trichloroacetic acid and/or l)y gel filtration im a talil)rated Sephadex C-200 column. "Cold" "carrier" sernm wa.s added to intestinal washings in order to aid precipitation or assist in identification (by SRID) of immunoglobulin-containing fractions. Immunoglohulins—quality The quality of radio-labelled iramunoglobulins was assessed by ( a ) determining the effects of tbe radio-labelling procedure on the vibriocidal activities of dog IgM and lgC2,, j, antibodies to V, cholerae. and. ( b ) comparing the decline in serum concentration of '-*'I-IgM with that of \ibriocidal IgM over 2-3 tlays following simultaneous intravenous injection of both marker popu]ati(jns (4 dogs). Determination of immimoelohulin transport Administration of lahelletl iiinmniofitobutiivi.
For 4 to 7 days before receipt of ^'^'I-
immunoglobulin (10 to 60 (iCi per dog), and thereafter until sacrificed, dogs ingested O-I to 0-2 g of KI daily. The labelled ininmnoglobulins were injected intravenously 9 to 17 days after surgery. Irrigation of Thiry-Vella loops. Loops were irrigated 46-72 h after administration of marker imiHuiuigloluilin in order to allow equililiration of marker immunoglobulin between vascular and extravastular Huids. Dogs to receive vibrioeidal IgM antibody were also irrigated and bled for "prc-immune" samples prior to injection of marker innnunoglobulin. Dogs were anaesthetised with sodium pentobarbital. The irrigating fluid was passed once through the loop. The irrigating solution eontained NaCI 90 mM. NaHCO., 50 niM and KCl 10 niM (Carpenter et al., 1968) and was introduced at appnix. 400 ml/h at a pressure not exceeding 15 cm of water. Washings were c'oUeeted on ice. Tbe first 150 ml of washings were discarded, following which 3 or 4 sueeessive 50 min eolleetions were made. These were designated W,, W.,. W^, and \V^. Bltxid was taken at tbe commencement and conclusion of tbe procedure. The dog was then killed and autopsy was performed immediately. Samples of the Thiry-Vella loop and of ileum and jejunum were placed in Bouin's Sxative for subsequent histological examination. Handling of intatiiial washings and sera. Washings were centrifuged in the cold at 8,000 g for 20 min. The supernatants were concentrated 20- to .3()-foId using negative pressure dialysis at 4° (Heddle. 197fi) and then divided into aliquots, some of which were frozen for later immunochemical analysis together with the corresponding serum samples. The pellets of "mucus" obtained by centrifuging washings were resuspended in 0-15 M saline and counted (for radioactivity). Cytological studies of intestinal tva-shiugs. Samples were obtained from resuspended washings prior to centrifugation and examined within 2-3 h ot collection. Red blood eells, total nucleated eells and polymorphonuclear leucocytes were counted using a haeniocytonieter. Polymorphonuelear leucocytes were identified using the peroxidase technique of R>t(jniaa (1962).
716
R. j . HEDDLE AND D . ROWLEY
Calculation nf "cleriraiKcs" ami "aiipcarame rates". Transport of marker fmTmmoglobulin was calculated as a clearance from scium into the intestinal lumen u-sing the following formula: — clearance = - r r — x ,—;;;- (ml m - ' h ^ ' ) . Ms L. I "Mw" and "Ms" represent the concentrations of marker irnmnnoglobulin in the concentrated intestinal washing and bI(M)d serum, respectively; "V". the xolun.ii' (nil) of concentrated intestinal washing; "L", the length (metres) of irrigated intesiiiie; and "T", the duration ( h ) of the irrigation. "Appearance rates" in the intestinal lumen of total IgG or IgM from all sources were calculated in the same units, as follows: — appearance rate =
„;
\
• ,_
(ml m - '
h-').
"C.w" and "Cs" represent the concentrations of total IgG or IgM in the intestinal washing and blood .serum. Clearances of '^^'I-immuno globulin were calculated in terms of radio-labelled immtmoglobulin removed specifically from washings or sera by the corresijonding immunoadsorbent. Shnilarly, vibriocidal titres were corrected for any contribution of pre-immune antibody.
RESULTS Autopsy results At autopsy the Thiry-Vella loops had the same appearance as the small bowel which remained in continuity. Sections of mid-portions of Thiry-Vella loops were similar in appearance to sections taken from jejunum and ileuin about 10 cm from the anastomosis sites. Cytology of intestinal washings Washings were examined microscopically for signs of bleeding and/or inflammation which might liave distorted the transport results. Dog red blood cells introduced d
THE SOURCE OE IgM AND IgG IN THE DOG INTESTINE by R. J. HEDDLE" AND D. ROWLEY (From the Department of Microbiology and Immunology, University of Adelaide, Adelaide 5001, South Australia.) (Accepted for publication October 10, 1978.) Summary. The source of IgM ant! IgGg., ,, in the dog intestine was investigated, using labflled dog iinmuriogloinilins and thionic Thiry-Vella loops of dog small intestine. Clearance.s of IgM and IgGg,, h from serum into intestinal secretions were similar, reflecting transfer of a greater ma.ss of IgGgy ^ than of IgM. Only 20% of the intestinal IgM derived from .serum, irrespective of the marker u.sed. Almiit one third of the total intestinal IgG. of all sub-classes and accrued from all sources, was seruni-ilerived IgG^,, i,. This indicates that serum is a major contributor to clog intestinal IgG^,, ,,.
INTRODUCTION Secretory IgA is quantitatively the dominant immunoglobulin in the lumen of the small intestine, but receut studies indicate that IgM and IgG antibodies or their fragments may play an important role in the intestine (Horsfall et al, 1978). The quantities of IgM and IgG transported from serum into the small intestine have been estimated in several studies. Wernet et al (1971) found that more (radio-labelled) rabbit IgG than rabbit IgM or secretory IgA was transported from serum into ligated loop.s of rabbit intestiue after cholera toxin stimnlation. Tlie proportions of each class of intestinal immunoglobulin transported from serum were not estimated. The major defect witli this study is that the permeability of the intestiual loops may have been increased (De and Chatterje, 1953). Recently, Cripps, Husband and Lascelles (1974) showed, in sheep, that the intestinal IgG2, most of the intestinal IgCi and an uncertain proportion of the intestinal IgM derived from serum. The contributions of the possible pathways by which immunoglobulins might reach the intestine are still not well understood, and this study was designed to examine the source of IgM and IgG in the dog small intestine using what was hoped would be a physiologically intact model. " Present address; Department of Medicine, Royal Adelaide Hospital, University of Adelaide, Adelaide 5001, South Australia.
7X4
R. J. HEDDLE AND D . ROWLEY MATERIALS AND M E T H O D S
Transport was detemiined using Thiry-Vella loops of dog small intestine which allowed easy acce.ss to small intestinal secretions and excluded other gastrointestinal secretions. Marker inmninoglobuiin was injected intra\enously. Transport was determined 2-3 days later by measuring the rate at which marker immunoglobulin entered the Thiry-Vella loop during irrigation. Three labelled dog immunoglobulins were used:— ^^^I-IgM. i''*iT-IgG.,.| ,, and IgM antibody capable of killing V. cholerae in the presence of complement. The latter was used to control whether the raitio-labelling procedures had changed the transport characteristics of the proteins a.s reported hy Tengstrom (1965) and Waldmann and Strober {1969). Recovery of "vibriocidar' IgM antibody indicated the presence of IgM antibody molecules with a high level of functional integrity. Dogs Adult mongrels, weighing 10-35 kg. were obtained from the Central Animal House of this Uni\ersil\\ Chronic Thiry-Vella loops were prepared in the middle third of the small liowel (Shields, Mulholland and Elmslie, 1966). The loops were irrigated every 2-3 days with isotonic "irrigating .solution" (description follows). Inimtitioglobulim—preparation and purity Dog IgG.j^i ,, and IgM were purified by published methods (Heddle and Rowley. 1975h). Double imuiunodittusion against rabbit anti-whole dog .serum and goat antisera monospecific for dog IgM (anti-n), IgG (anti-y) and (secretory) IgA (anti-a) (Heddie and Rowley. 1975a; Heddle. 1976) revealed no impurities. Immunoglobulins were labelled with carrier-free ' ^U (Amersham, Buckinghamshire, England) using the chloramine-T technicjue of McConahey and Dixon (1966). Radiolabelled immunoglobulins were separated from free radio-iodine by gel filtration (Sephadex G-25). Resulting specific activities were approx. 10 nCi/mg. Serum was obtained from a dog wbich bad been immunised with washed live V. cholerae 569B (Neoh and Rowley. 1970). "Vibriocidal" IgM antibody was obtained by gel filtration of this serum (Heddle, 1976). The same procedure demonstrated that the vibriocidal activity in the immune serum was associated with IgM antibody. Immunoglobulins—measurement and efiaracterisation in sera and intestinal washings Single radial immunodiffusion (SRW). Total IgM, IgG and IgA were measured by SRID (Mancini. Carbonara and Heremans. 1965) using the respective monospecific antisera and appropriate standards (Heddle, 1976). The cibriocidal assay was based on that described by Neoh and Rowley (1970). The vibriocidal acti\ity of intestinal washings was characterised by incubating (.'37°, lh) washings before assay with equal voluuies of (a) 0-2 M mercaptoethanol (freshly diluted in phosphate buffered saline (PBS). pH 7-4). (b) V. cholerae 569B lipopolysaceharide (Steele, Chaicumpa and Rowley. 1974) at 40-100 ng/ml in PBS. or (c) PBS. Samples were dialysed against PBS and titrated in duplicate using doubling dilutions. Coiitiiing of radioactivity was performed in a Packard Autogamma Spectrometer. Samples were counted to a niinimuni of 1,000 counts greater than background and counts were corrected for decay of radioacti\ity. Counting errors were estimated using standard formulae (Nuclear-Chicago Corporation. 1967). Immunopredpitation of radioactivity was achieved using monospecific goat anti-dog sera which had been prepared and linked covalently to agarose beads as described (Heddle and Rowley, 1975 a and b). Duplicate aliquots of samples were e.\posed (room temperature, 30 min) to either an excess of the appropriate anti-dog heavy chain immunoadsorbent or an
TRANSPORT OF DOG IMMUNOGLORULINS
715
equal quantity of a beterologoiis iniuumoaclsorbent. Followinjj; this, the inimunoadsorbents were wasbed live times in ice-cold PBS (pH 7-1) and counted. If "I" represents tbe counts which were present in the sample. "S" the counts on tbe washed homologous ininuuioadsorbent, and "C" the eounts on the washed heterologous imnuinoadsorbent, tbe proportion of the available eoimts adsorbed specificaliy was taken as S-C divided by I-C Binding to tbe heterologous immunoadsorbent was assumed to be representative of nonspecific binding to the homologous immunoadsorbeut. The effect of counting errors on the accuracy of this calculation was estimated using published formulae (Nuclear-Chicago Corporation. 1967; Arinitage, 1971). Trichloruacetic aciil preparation and gel filtration. Some radioactivf samples were analysed further by precipitation with an eqna! volume of cotd 2O'-6 trichloroacetic acid and/or l)y gel filtration im a talil)rated Sephadex C-200 column. "Cold" "carrier" sernm wa.s added to intestinal washings in order to aid precipitation or assist in identification (by SRID) of immunoglobulin-containing fractions. Immunoglohulins—quality The quality of radio-labelled iramunoglobulins was assessed by ( a ) determining the effects of tbe radio-labelling procedure on the vibriocidal activities of dog IgM and lgC2,, j, antibodies to V, cholerae. and. ( b ) comparing the decline in serum concentration of '-*'I-IgM with that of \ibriocidal IgM over 2-3 tlays following simultaneous intravenous injection of both marker popu]ati(jns (4 dogs). Determination of immimoelohulin transport Administration of lahelletl iiinmniofitobutiivi.
For 4 to 7 days before receipt of ^'^'I-
immunoglobulin (10 to 60 (iCi per dog), and thereafter until sacrificed, dogs ingested O-I to 0-2 g of KI daily. The labelled ininmnoglobulins were injected intravenously 9 to 17 days after surgery. Irrigation of Thiry-Vella loops. Loops were irrigated 46-72 h after administration of marker imiHuiuigloluilin in order to allow equililiration of marker immunoglobulin between vascular and extravastular Huids. Dogs to receive vibrioeidal IgM antibody were also irrigated and bled for "prc-immune" samples prior to injection of marker innnunoglobulin. Dogs were anaesthetised with sodium pentobarbital. The irrigating fluid was passed once through the loop. The irrigating solution eontained NaCI 90 mM. NaHCO., 50 niM and KCl 10 niM (Carpenter et al., 1968) and was introduced at appnix. 400 ml/h at a pressure not exceeding 15 cm of water. Washings were c'oUeeted on ice. Tbe first 150 ml of washings were discarded, following which 3 or 4 sueeessive 50 min eolleetions were made. These were designated W,, W.,. W^, and \V^. Bltxid was taken at tbe commencement and conclusion of tbe procedure. The dog was then killed and autopsy was performed immediately. Samples of the Thiry-Vella loop and of ileum and jejunum were placed in Bouin's Sxative for subsequent histological examination. Handling of intatiiial washings and sera. Washings were centrifuged in the cold at 8,000 g for 20 min. The supernatants were concentrated 20- to .3()-foId using negative pressure dialysis at 4° (Heddle. 197fi) and then divided into aliquots, some of which were frozen for later immunochemical analysis together with the corresponding serum samples. The pellets of "mucus" obtained by centrifuging washings were resuspended in 0-15 M saline and counted (for radioactivity). Cytological studies of intestinal tva-shiugs. Samples were obtained from resuspended washings prior to centrifugation and examined within 2-3 h ot collection. Red blood eells, total nucleated eells and polymorphonuclear leucocytes were counted using a haeniocytonieter. Polymorphonuelear leucocytes were identified using the peroxidase technique of R>t(jniaa (1962).
716
R. j . HEDDLE AND D . ROWLEY
Calculation nf "cleriraiKcs" ami "aiipcarame rates". Transport of marker fmTmmoglobulin was calculated as a clearance from scium into the intestinal lumen u-sing the following formula: — clearance = - r r — x ,—;;;- (ml m - ' h ^ ' ) . Ms L. I "Mw" and "Ms" represent the concentrations of marker irnmnnoglobulin in the concentrated intestinal washing and bI(M)d serum, respectively; "V". the xolun.ii' (nil) of concentrated intestinal washing; "L", the length (metres) of irrigated intesiiiie; and "T", the duration ( h ) of the irrigation. "Appearance rates" in the intestinal lumen of total IgG or IgM from all sources were calculated in the same units, as follows: — appearance rate =
„;
\
• ,_
(ml m - '
h-').
"C.w" and "Cs" represent the concentrations of total IgG or IgM in the intestinal washing and blood .serum. Clearances of '^^'I-immuno globulin were calculated in terms of radio-labelled immtmoglobulin removed specifically from washings or sera by the corresijonding immunoadsorbent. Shnilarly, vibriocidal titres were corrected for any contribution of pre-immune antibody.
RESULTS Autopsy results At autopsy the Thiry-Vella loops had the same appearance as the small bowel which remained in continuity. Sections of mid-portions of Thiry-Vella loops were similar in appearance to sections taken from jejunum and ileuin about 10 cm from the anastomosis sites. Cytology of intestinal washings Washings were examined microscopically for signs of bleeding and/or inflammation which might liave distorted the transport results. Dog red blood cells introduced d
