CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2005;3:687 694 Effect of Pretretment of Food Gluten With Prolyl Endopeptidse on Gluten-Induced Mlbsorption in Celic Sprue GAIL G. PYLE,* BRIAN PAASO, BARBARA E. ANDERSON, DIANE D. ALLEN,* THOMAS MARTI,* QING LI,* MATTHEW SIEGEL, CHAITAN KHOSLA,* nd GARY M. GRAY* *Celic Sprue Reserch Foundtion, Plo Alto; Helth Cre Division nd Reserch Institute of Plo Alto Medicl Foundtion, Plo Alto; nd Deprtment of Chemicl Engineering, Stnford University, Stnford, Cliforni Bckground & Aims: We sought to determine whether prolyl endopeptidse (PEP) tretment of food gluten would obvite the intestinl dysfunction produced by smll mounts of dietry gluten supplement in ptients with celic sprue. Methods: Twenty symptomtic ptients with histologiclly proven celic sprue completed rndomized, double-blind, cross-over study involving two 14-dy stges. Ech ptient consumed low dose of gluten supplement dily (5 g; equivlent to 1 slice of bred) in 1 stge nd gluten pretreted with PEP in the other stge. Ptients completed dily symptom questionnire nd D-xylose urine excretion nd 72-hour quntittive fecl ft were monitored before nd fter ech stge. Results: Despite clinicl remission t bseline, 40% of ptients hd t lest 1 bnorml celic ntibody, 20% hd n bnorml urine xylose, nd 63% hd n bnorml fecl ft test result. There ws no difference in symptoms s function of the type of gluten consumed. In response to gluten not treted with PEP, n pprecible proportion of ptients developed mlbsorption of ft (7 of 17, 41%) or xylose (8 of 14, 57%). When the gluten ws pretreted with PEP, ft mlbsorption ws voided in 5 of 7 nd xylose mlbsorption in 4 of 8 of these sme ptients. Conclusions: A significnt proportion of symptomtic ptients with celic sprue hve bnorml celic ntibodies nd ft or crbohydrte mlbsorption. Pretretment of gluten with PEP voided the development of ft or crbohydrte mlbsorption in the mjority of those ptients who developed ft or crbohydrte mlbsorption fter 2-week gluten chllenge. Celic sprue (celic sprue lso hs been clled celic disese, gluten-sensitive enteropthy, nd nontropicl sprue) is gluten-induced disese of the smll intestine chrcterized by lymphocytic inflmmtion with distortion of the mucosl rchitecture, extensive loss of the surfce re, nd consequent mlbsorption of nutrients nd vitmins. In the clssic disese, symptoms of bdominl bloting, distention, nd discomfort re ssocited with the output of bulky stools nd consequent weight loss. However, substntil proportion of ptients experience only miniml nonspecific symptoms of postprndil fullness, bloting, nd mild middy ftigue, despite the presence of pprecible intestinl dmge. Celic sprue ppers to be cused by sequence of events: (1) the trnsglutminse-ctlyzed conversion of selected glutmine to glutmic cid residues in poorly digested gluten peptides from whet, brley, nd rye, which thereby enhnces the following: (2) their subsequent binding to humn leukocyte ntigen DQ2 (or DQ8) sites on mucosl ntigen-presenting cells; nd (3) the consequent lymphocytic prolifertion nd cscde of inflmmtory nd secretory events leding to intestinl surfce flttening. 1,2 Although exclusion of gluten s dietry source is effective in reversing the intestinl dmge in the mjority of ptients, lifelong gluten-free diet is difficult to mintin becuse gluten is not only one of the most widespred components of the humn diet, but it is lso n unlbeled ingredient in pckged foods. Other therpies hve not been estblished, but the first line of ttck my be supplementl therpy with peptidse hving high specificity for the digestive-resistnt, proline-rich gluten peptides, thereby eliminting them from intercting with the smll intestine. The prolyl endopeptidses (PEPs) re puttive cndidte enzymes tht might be dministered t the time of mel to be relesed or ctivted in the upper intestinl lumen where they could complement the pncretic proteses by further processing the toxic gluten peptides in the intestinl lumen, thereby Abbrevitions used in this pper: ANOVA, nlysis of vrince; Ig, immunoglobulin; PEP, prolyl endopeptidse; PTC-gluten, pepsin, trypsin, nd chymotrypsin lone. 2005 by the Americn Gstroenterologicl Assocition 1542-3565/05/$30.00 PII: 10.1053/S1542-3565(05)00366-6
688 PYLE ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 7 pre-empting their interction with the intestinl surfce. 3,4 The tretment of proteolyzed gluten with PEP hs been shown to decrese substntilly the number of potentilly immunostimultory peptides nd to reduce the immunogenicity of the resulting mteril towrd intestinl polyclonl T-cell lines from celic sprue ptients. 5 In ddition, in preliminry studies (see our compnion report 6 in this issue), we were ble to estblish tht low-dose, short-term (5 10 g/dy for 14 dys) gluten orl supplement induces ft nd crbohydrte mlbsorption when given to symptomtic celic sprue ptients on n otherwise gluten-free diet. This protocol now hs been modified to exmine the effect of gluten pretreted with pepsin, trypsin, nd chymotrypsin lone (PTC-gluten) s compred with PTC-gluten dditionlly treted with PEP (PTCgluten PEP). In this study, celic ptients, mintining n otherwise gluten-free diet, ingested supplementof5gofglutenorglutentreted with PEP in n ornge- or lemon-juice vehicle for 14 dys in double-blind, cross-over mnner. The course of ptient symptoms nd the response of intestinl bsorptive prmeters (72-hour fecl ft nd 5-hour xylose urinry excretion) were monitored before nd fter ech segment of the study. Methods Ptients nd Study Design This study ws pproved by the institutionl review bord of the Plo Alto Medicl Foundtion in Plo Alto, Cliforni. All prticipnts in the study were counseled regrding risks nd signed n informed consent document. Twenty-two symptomtic ptients (7 men, 15 women; ge, 21 78 y; men ge, 49 y) who were mintining glutenexclusion diet were recruited. The time since dignosis for the 20 ptients rnged from 3 months to 18 yers (men, 6 y). For 13 ptients, the dignosis ws mde more thn 2 yers before this study (rnge, 2.8 18 y; men, 9.1 y); 9 ptients hd the dignosis for 2 yers or less (rnge,.25 2 y; men, 1.1 y). Two ptients dropped out of the study, one ner the end of the first stge (ptient I, becuse of n cute respirtory illness) nd the other before the second stge (ptient R, becuse of conflicting personl commitments). Ptients provided copies of their initil smll-intestinl biopsy exmintion pthology report to verify dignosis of celic sprue. At dignosis, the 14 ptients tested were found to hve t lest 1 elevted ntibody (57%, 43%, nd 50% hd bnorml trnsglutminse, endomysil, or ntiglidin ntibody levels, respectively). An entry questionnire concerning the medicl history, dherence to the gluten-free diet, nd current symptoms ws completed. A double-blind, cross-over design ws used for ech symptomtic prticipnt in 2 chllenge stges, in rndomized order. In 1 stge, ech ptient consumed low dose of gluten supplement (5 g; equivlent to 1 slice of bred) tht ws predigested with PTC-gluten; in the other stge, ech ptient consumed PTC-gluten PEP. Ech stge consisted of 14 dys, seprted by wshout period of t lest 6 weeks. In the first stge, ptients were rndomized to consume either the control ornge juice mixture contining 5 g of PTC-gluten, or the PEP ornge juice mixture (contining 5gofPTC-gluten PEP) dily for 14 dys s supplement to ptients usul gluten-free diet. After being exposed to the PTC-gluten, the PEP ws inctivted completely through heting process before being incorported into the ornge juice mixture. After the wshout period, the ptients switched in the second stge to consuming the other type of ornge juice mixture dily for 14 dys. Two of the 20 ptients (ptients U nd V) prticipted in our prior gluten chllenge study. 6 Ptients in tht study consumed either 5 or 10 g of gluten (in n identicl orngejuice mixture to the one used in this study) dily for 21 dys, with the 72-hour fecl ft test nd the 25-g, 5-hour urine xylose test performed t the beginning nd t the end of the second week. The 2 ptients described erlier hd both been ssigned to the 10-g group. Becuse there ws no sttisticl difference in ny of the test results between the ptients who received 5gofgluten per dy nd those who received 10 g of gluten per dy, 6 we used the 2 ptients dt from the prior study s their control stge dt for this study rther thn hving them repet the control stge. Becuse these ptients were wre tht they were consuming PTC-gluten PEP, only the bsorptive tests nd not the symptom questionnire dt were included for sttisticl nlysis. Studied Vribles During ech stge of the study, ptients completed dily questionnire concerning symptoms nd dherence to the gluten-free diet. The symptom list ws chnged slightly from our initil gluten-chllenge study by the elimintion of 2 items (skin rshes nd sore/red tongue) tht were not experienced by ny ptient in tht preliminry study. 6 Ptients documented the presence or bsence of the other 11 symptoms (detiled in our compnion report 6 in this issue) by use of 0 3 ordinl scle (none, mild, moderte, nd severe), yielding mximum possible dily symptom score of 33. Ptients lso indicted their perceived dherence to gluten-free diet by use of 0 3 scle (with 0, 1, 2, nd 3 representing no ingestion of gluten; mild, moderte, nd severe intke of gluten, respectively). Lbortory testing ws performed s detiled in the compnion report 6 in this issue with the following exceptions: the celic ntibody pnel ws performed only t bseline (within 2 weeks of the strt of the study), nd the 25-g D-xylose urine excretion test nd the 72-hour quntittive fecl ft test were performed t the beginning nd t the end of ech of the 2 stges.
July 2005 PEPTIDASE TREATED GLUTEN IS NONTOXIC 689 Preprtion of Pepsin, Trypsin, nd Chymotrypsin (PTC) Gluten nd Prolyl Endopeptidse Treted PTC-Gluten Commercilly vilble whet gluten ws incorported into n ornge juice mixture s detiled in our compnion report 6 in this issue. Frozen ornge juice concentrte (Minute Mid, thwed, 2 oz/dose; Minute Mid, Houston; TX) nd lemon juice concentrte (Minute Mid,.5 oz/dose) were dded to the PTC-proteolyzed gluten solution, mixed, nd trnsferred to plstic beverge continers (8 oz/5 g gluten). The continers were stored in 20 C freezer until the dy before use. Anlysis reveled tht the gluten digests were stble by high-performnce liquid chromtogrphy nlysis for t lest 90 dys. The PEP gene ws mplified from the genomic DNA (F meningosepticum: ATCC 13253), cloned into pet28b plsmid (Novgen, Mnsss, VA), nd purified by nickel ffinity chromtogrphy s detiled by Shn et l. 7 The finl pure mteril hd specific ctivity of 37 mmol/l 1 s 1 (40 U/mg protein) with succinyl-l-pro-p-nitronilide s substrte. The pure PEP ws dded t rtio of 200 U to 1gofthePTC-digested gluten substrte, incubted t 37 C for 1 hour, nd then het dectivted t 95 C for 1 hour. 5 Anlysis of the PTC-gluten nd PTC-gluten PEP ws performed on ech btch of gluten preprtion by C18 high-performnce liquid chromtogrphy, s detiled previously. 5 The PEP tretment of the PTC-gluten produced mrked reduction of the glidin peptide end products nd resulted in relted increse in the severl smller peptide products tht re no longer cpble of stimulting T cells from celic ptients (see Mrti et l 5 for high-performnce liquid chromtogrphy pttern, peptide structure, nd nlysis of T-cell repliction). Sttisticl Anlysis Dily symptoms, urine xylose, nd fecl ft output ll were expected to worsen with gluten chllenge (PTC-gluten). If PEP tretment eliminted the negtive effect of gluten, vlues for ll vribles were expected to remin unchnged from norml bseline vlues fter consumption of PTC-gluten PEP. Repeted-mesures nlysis of vrince (ANOVA) tests (with set t.05) were used to determine ptients differences in xylose nd fecl ft cross pre-pep, post-pep, pre-gluten, nd post-gluten mesurement times. Bonferroni t tests 8 (with n djusted of.012 for 4 comprisons of interest) were used for the plnned multiple comprisons between prestge nd poststge mesures fter ANOVA testing, with 1-tiled tests becuse the hypothesis nticipted chnge in only 1 direction. The order effect (hving the PEP stge or gluten chllenge stge first) ws hypothesized to be zero, nd ws tested with 2-wy ANOVA nd with t tests compring pre-pep nd pre-gluten bseline mesures cross ptients for ech vrible. 2 nlyses were used to compre the bseline vlues of ptients who hd greter thn or less thn 2 yers since their dignosis, nd those who reported strict vs not-so-strict dherence to gluten-free diet in the 2 weeks before the study. Results Physicl Findings The mjority of ptients hd no significnt physicl findings. At the initil physicl exmintion, 4 ptients hd mild subjective bdominl tenderness. During the study, 1 ptient developed mild mculoppulr rsh ttributed to known mild skin rection, possibly cused by ornge juice, which did not interfere with the ptient s bility to complete the study. Questionnire Dietry dherence. Prticipnts were sked to mintin strict gluten-free diet during the study. There ws no significnt difference in dietry dherence between the 2 stges for ny of the 20 ptients. Eight of the ptients (40%) reported no perceived episodes of gluten contmintion in their food intke during either stge. The mjority of the other ptients perceived ingesting smll mounts of gluten-contining foods (mild contmintion) for 1 3 dys in one or both stges. Two ptients (10%) reported contmintion on t lest hlf of the dys of ech stge. Both of these ptients hd 2 bnorml bseline tests. In ddition, 5 of the 20 ptients (including the 2 ptients described erlier) perceived tht they hd mild gluten contmintion sometime during the 2 weeks before entry into the study, wheres the mjority of ptients (15 of 20; 75%) reported tht they hd mintined strict gluten-free diet during tht time frme. There ws no difference in xylose, fecl ft, or ntibody titers t bseline between those who reported strict dherence nd those who reported some gluten contmintion in the 2 weeks before the study. Dily Symptoms Dily symptom scores rnged from 0 to 22, with mximum possible score of 33. Most ptients hd reltively few symptoms during either the control or PEP tretment stge. The verge totl stge symptom score (the sum of the 14 dily symptom scores for tht stge out of mximum of 462) for the 20 ptients who completed the study ws 22 (rnge, 0 71) for the control (gluten, no PEP) stge, nd 23 (rnge, 4 82) for the gluten PEP stge, which ws not significnt difference. Ptient R, who dropped out of the study fter completing the control stge, hd the highest totl stge symptom score of 89.
690 PYLE ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 7 Tble 1. Bseline Lbortory Vlues for Ptients on Gluten-Free Diet for 2 Yers Ptient Test A D E G I L O Q T Ttg IgG (norml, 20) 8 16 6 7 8 8 5 3 5 Ttg IgA (norml, 20) 161 19 6 19 13 200 47 15 9 AGA IgA (norml, 20) 21 19 7 10 11 91 12 8 11 Fecl ft 12 5.1 1.9 22 4.4 13 7.6 25 (norml, 7 g/24 h) Urine xylose 6.7 6.2 6.0 5.9 4.2 2.2 6.2 (norml, 4 g/5 h) Yers since dignosis.25 2.67 1 2.5 2.33 1.33 NOTE. Abnorml vlues indicted in bold. Ttg IgG, tissue trnsglutminse IgA ntibody; Ttg IgA, tissue trnsglutminse IgA ntibody; AGA IgA, nti-glidin IgA ntibody; fecl ft, 72-hour quntittive fecl ft nlysis; urine xylose, 5-hour urine xylose nlysis. Indequte or lost smple. Bseline Lbortory Vlues None of the 20 ptients who completed the study ws immunoglobulin (Ig)A deficient. Bseline nlyses re shown in Tbles 1 nd 2. Eight ptients (40%) hd t lest 1 bnorml ntibody level t bseline (Tbles 1 nd 2, shded vlues). Five of the 20 ptients (25%) hd missing bseline xylose urine test results owing to lbortory error. Of the remining 15 ptients, 3 (20%) hd n bnorml urine xylose vlue t bseline. One of the 20 ptients hd missing bseline fecl ft result, gin owing to lbortory error. Twelve of the remining 19 ptients (63%) hd n bnorml bseline fecl ft vlue, but only 4 of those 12 ptients (33%) hd bnorml ntibody levels. There ws no difference in bseline lbortory vlues between those who hd been dignosed more thn 2 yers before the study, nd those who hd been dignosed 2 yers or less before the strt of the study (Tbles 1 nd 2). One of the 2 ptients who dropped out of the study hd norml bseline vlues. The other ptient hd norml celic ntibodies nd norml fecl ft vlue (xylose vlues were missing). Gluten Chllenge nd Prolyl Endopeptidse To determine whether PEP pretretment of gluten ws effective in voiding mlbsorptive response, we identified those ptients who hd positive mlbsorptive response to the gluten chllenge bsed on fecl ft or xylose testing. For those ptients, the puttive voidnce of mlbsorptive response when chllenged with the PTC-gluten PEP then could be considered. Criteri for positive gluten chllenge. A positive gluten chllenge ws bsed on mlbsorption when the prestge nd poststge xylose vlues met either one of the following criteri: (1) prestge bsorption (xylose or fecl ft) ws norml nd the poststge vlue ws bnorml, or (2) the bseline bsorptive tests for xylose or fecl ft were bnorml, nd t lest 25% decrese in the urine xylose excretion or increse in the fecl ft output from prestge to poststge. The number of ptients s func- Tble 2. Bseline Lbortory Vlues for Ptients on Gluten-Free Diet for 2 Yers Ptient Test B C F H J K M N P R S U V Ttg IgG (norml, 20) 9 8 7 33 10 16 5 3 5 5 5 10 35 Ttg IgA (norml, 20) 16 43 5 10 119 26 8 11 3 14 8 7 10 AGA IgA (norml, 20) 14 11 8 7 40 23 9 11 8 16 17 9 11 Fecl ft 10 14 4.5 16 5.3 3.4 8.2 11 11 6.6 15 6 3.9 (norml, 7 g/24 h) Urine xylose 4.8 5.3 2.2 5.5 3.0 4.0 6.1 9.9 5.6 (norml, 4 g/5 h) Yers since dignosis 3.25 18 8 10 8 3.17 6.67 11 11 4 17 14.8 2.75 NOTE. Abnorml vlues indicted in bold. Ttg IgG, tissue trnsglutminse IgA ntibody; Ttg IgA, tissue trnsglutminse IgA ntibody; AGA IgA, nti-glidin IgA ntibody; fecl ft, 72-hour quntittive fecl ft nlysis; urine xylose, 5-hour urine xylose nlysis. Indequte or lost smple.
July 2005 PEPTIDASE TREATED GLUTEN IS NONTOXIC 691 Figure 1. Numbers of symptomtic celic sprue ptients in the gluten nd gluten-pep chllenge s function of the (A) xylose urine excretion test nd the (B) quntittive fecl ft excretion test. tion of mlbsorptive tests nd response to the gluten chllenge is shown in Figure 1. D-Xylose 5-Hour Urine Test Fourteen ptients hd full set of urine xylose dt to nlyze (Figure 1). In 5 of the 6 excluded ptients with n invlid xylose test (ptients B, F, G, K, nd N), the bseline urine smple volume ws not recorded. The sixth excluded ptient (ptient L) hd insufficient urinry output (35 ml for 5-hour collection). Repeted-mesures ANOVA reveled significnt differences cross times for these 14 ptients (P.02), so we proceeded with the comprtive sttisticl nlysis with the pired t test. As shown in Tbles 3 nd 4, the pired t tests reveled significnt reductions between prestge PTCgluten nd poststge PTC-gluten xylose excretion (for ll 14 ptients, P.004; for 8 with positive gluten chllenge (s defined erlier under Criteri for positive gluten chllenge), P.001; nd for 4 with strict criteri of norml-to-bnorml xylose response to PTC-gluten, P.02). In contrst, there ws no chnge in xylose excretion for the pre-stge PTC-gluten PEP vs poststge PTCgluten PEP (ll 14 ptients, P.13; 8 with positive gluten chllenge, P.14, for 4 with norml prestge to bnorml poststge xylose P.38; Tbles 3 nd 4). PEP tretment of the PTC-gluten ws effective in obviting the development of xylose mlbsorption in the other stge in 4 of the 8 gluten-responsive ptients (50%). In those ptients fitting the stricter criteri for gluten response (Tble 4, ptients E, M, O, nd V), 3 of the 4 mintined norml xylose bsorption with the PEP-treted gluten (Tble 4, ptients M, O, nd V). 72-Hour Quntittive Fecl Ft Anlysis The stool for fecl ft nlysis ws lost in 1 of the 20 ptients (ptient O), nd 2 ptients (ptients F nd L) hd n incomplete smple collection ( 150 g/72 h) for 1 of the 4 stool collections. Hence, these ptients were excluded for the fecl ft nlysis. When nlyzing subgroups of ptients bsed on the criteri described erlier, 7 of the 17 ptients (41%) with complete dt hd positive mlbsorptive response to the gluten chllenge bsed on the fecl ft test (Tbles 3 nd 5). Following repeted-mesures ANOVA (time difference, P.05), the pired t test reveled significnt increse in ft excretion by the end of the 2-week gluten chllenge (prestge vs poststge PTC-gluten comprisons, Tble 5, P.0003). In these 7 gluten-responsive ptients, ingestion of PTC-gluten PEP voided the development of ft mlbsorption in 5 (71%; Tble 5, ptients B, M, P, U, nd V). In contrst, there ws no increse in their ft excretion when compring prestge vs poststge in the PTC-gluten PEP segment (P.40). For the 4 ptients (Tble 5, ptients B, P, U, nd V) with the strict criteri of norml bseline fecl ft vlue who developed stetorrhe in the PTC-gluten stge (P.006), ingestion of PTC-gluten PEP, voided stetorrhe in ll 4 (Tble 5; P.11). Order Effect Whether ptients consumed the PTC-gluten or the PTC-gluten PEP in the first stge did not seem to ffect the results. ANOVA testing reveled no order effect cross the totl sum of ptients. In ddition, no difference in bseline chrcteristics ws found between Tble 3. Gluten Chllenge PEP Pretretment: Urine Xylose nd Fecl Ft Ptients with full xylose dt (N 14) Ptients with full fecl ft dt (N 17) Positive gluten chllenge 8/14 57% 7/17 41% PEP effective 4/8 50% 5/7 71% NOTE. See text under Results for D-xylose 5-hour Urine Test nd 72-hour fecl ft nlysis for relted P vlues. Positive chllenge is chnge from norml to bnorml bsorptive test, or, if bnorml t bseline, worsening by t lest 25%.
692 PYLE ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 7 Tble 4. Urine Xylose for Ptients With Positive Xylose Gluten Chllenge Ptient No PEP (P.001) PEP (P.14) Prestge Poststge Prestge Poststge D 9.4 4.4 6.2 7.2 E 6.0 1.3 3.4 6.1 J 2.2.8 4.3 2.5 M 5.5 3.2 7.8 3.6 O 4.2 3.1 4.3 4.1 Q 2.4.8 2.2 1.6 U 9.9 4.2 7.3 1.4 V 5.6 3.2 6.2 6.0 NOTE. The norml rnge for urine xylose is 4 g/5 h. Positive chllenge is chnge from norml to bnorml xylose or reduction by t lest 25%. The pired t test reveled reduction of xylose excretion in the PTC-gluten (no PEP) stge (P.001) but no decresed xylose excretion in the PTC-gluten PEP stge (PEP, P.11). If only those with norml to bnorml vlues during the no PEP stge re defined s hving the gluten response (ptients E, M, O, nd V), then xylose mlbsorption developed for no PEP stge (P.02) but not for the PEP stge (P.38). those ptients consuming the PTC-gluten in the first stge nd those consuming PTC-gluten PEP in the first stge. Discussion Our preliminry study on the bsorptive cpcity of symptomtic celic sprue ptients reveled (1) mlbsorption of ft nd crbohydrte in minority of ptients in the bsence of ny significnt intestinl symptoms, nd (2) low-dose short-term gluten supplementtion (5 or 10 g of gluten/dy for 14 dys) induced symptoms nd mlbsorption in the mjority of these celic ptients. 6 Becuse the prticipnts in the preliminry study were wre they were consuming gluten supplement in tht tril, the ugmented symptoms might hve been nticipted s plcebo effect. In this double-blind study, symptoms were not distinguishble during the control (PTC-gluten) vs the tretment (PTCgluten PEP) stge, even though there ws significnt functionl mlbsorptive response (see Tbles 3 5). We sought to determine whether PEP-treted gluten would obvite the intestinl dysfunction produced by smll mounts of dietry gluten supplement in ptients with celic sprue. PEP-treted gluten hs been shown to metbolize mny toxic glidin peptides of whet gluten nd to reduce mrkedly their cpcity to induce T-cell repliction in vitro. 5 Indeed, this study indictes tht peptidse inctivtion of the glidin peptides in food gluten does circumvent the functionl mlbsorptive effects of the digestive-resistnt gluten peptides in ptients with celic sprue. A confounding fctor in this study ws the reltively high prevlence of bnorml bsorptive function in those with celic sprue who were symptomtic nd considered to be in remission both by the ptients themselves nd by their physicins. In ddition, only bout hlf of those chllenged with the 5 g of food gluten developed mlbsorptive tests. Therefore, we chose to perform 3 levels of nlysis (see Figure 1): (1) ll ptients with vlid xylose or fecl ft tests; (2) ptients with vlid tests who developed mlbsorptive response s defined erlier to PTC-gluten lone; nd (3) ptients from group 2 who hd the strict criteri of norml bsorptive tests tht converted to bnorml in response to PTC-gluten lone. As detiled in the Results section nd Tbles 4 nd 5, ll 3 levels of nlysis reveled highly significnt reduction of the prevlence of mlbsorption of both xylose nd ft in response to the low-dose gluten chllenge. This substntive finding encourges us tht the PEP is suitble puttive supplementry digestive enzyme for glidin peptide processing. Future studies should verify whether it is possible to deliver the PEP effectively to the duodenum where it my complement the pncretic proteses in ptients with celic sprue. The quntittive fecl ft test hs been known over the yers to be the most sensitive test for documenting mlbsorption, wheres the less-sensitive 5-hour urine xylose test, when positive, loclizes the condition to the smll intestine rther thn to defect in exocrine pncretic function. Although n upper-gstrointestinl endoscopy nd distl duodenl biopsy exmintion could not be justified in this initil study of helthy volunteers with the disese, the estimtion of bsorptive cpcity of Tble 5. Fecl Fts of Ptients With Positive Fecl Ft Gluten Chllenge Ptient No PEP (P.003) PEP (P.40) Prestge Poststge Prestge Poststge B 4.3 12 10 5.7 M 8.2 11 7.7 4.6 P 4.9 12 11 5.8 S 15 24 8.2 25 T 16 22 25 15 U 6 11 7.8 5.6 V 3.9 7 3.4 5.4 NOTE. The norml rnge for the fecl ft test is 7 g/24 h. Positive gluten chllenge is defined s chnge from norml to bnorml, or, if bnorml t bseline, n increse by 25% in the no PEP stge. The incresed ft mlbsorption with PTC-gluten (no PEP, P.003) ws voided by treting the gluten with PEP (PEP, P.40). If only those with norml to bnorml chnge in response to gluten lone (no PEP) re considered (ptients B, P, U, V), then the pired t test still revels mlbsorptive response with gluten lone (P.006) but no mlbsorptive chnge occurred in the PEP stge (P.11).
July 2005 PEPTIDASE TREATED GLUTEN IS NONTOXIC 693 ft nd crbohydrte, while somewht cumbersome, involved essentilly no risk to the celic ptient. Other prmeters of intestinl function, including urinry excretion of nitric oxide, 9 ugmented intestinl permebility of nontrnsportble oligoscchrides, 10 nd incresed interleukin-2 receptor concentrtions in plsm 11,12 lso were considered. We initilly did mesure interleukin-2 receptor levels, but found tht there ws pprecible nd seemingly rndom vrition in vlues in individul ptients nd between ptients, with reltively poor correltion to the better-estblished bsorptive prmeters in these gluten vs gluten PEP chllenge experiments. No sttisticlly significnt difference for soluble plsm interleukin-2 receptor ws found for gluten lone vs gluten PEP stges of this study (P.58). In the design of gluten-chllenge study, it is importnt tht the length of the study be reltively short, the gluten supplement nd the testing be tolerble to the ptients, nd the schedule of feeding gluten ( pretretment with PEP) be sufficiently prcticl to llow preprtion of sufficient quntities of the peptidse. Although the initil bckground experiments (detiled in our compnion report 6 in this issue) indicted tht the reltively low dose of 5 g, equivlent to smll slice of bred, ws s effective s 10 g in inititing mlbsorption uniformly fter dily ingestion for 2 weeks, 6 not ll of the celic sprue ptients in this blinded study responded to this low dose with ft nd crbohydrte mlbsorption (see Tbles 3, 4, nd 5). Now tht it hs been shown tht mlbsorption is very common in symptomtic celic sprue volunteers, documenttion of norml ft nd crbohydrte bsorption should be correlted with the structurl integrity of the intestine bsed on distl duodenl biopsy specimens obtined t upper-gstrointestinl endoscopy. Another reltively noninvsive test of intestinl structure of more distl levels of jejunum nd ileum, the rdiotelemetry videocpsule lso my provide incrementl informtion for ssessment of overll smll intestinl structure. 13 In ddition, we would now fvor dding n oligoscchride permebility test s n dditionl ssessment of intestinl integrity. Celic serum ntibody titers (glidin, endomysil, trnsglutminse) correlte closely with the histologic findings in untreted celic sprue 14 16 nd frequently re used to monitor dherence to gluten-exclusion diet. 17 20 In our preliminry compnion study, 6 we found tht 21 dys of low-dose (5 10 g/dy) orl gluten supplementtion does not yield n increse in these ntibody titers to the bnorml rnge. 6 The current study of symptom-free individuls with celic sprue reveled tht 8 ptients (Tbles 1 nd 2, ptients A, C, H, J, K, L, O, V) hd incresed trnsglutminse titers t bseline, 6 of these being IgA-type ntibodies nd 2 of the IgG type. Five of these ptients lso hd mlbsorption, nd 4 ptients were positive for glidin IgA. Prticulrly striking ws the fct tht ll celic ntibodies were negtive in 8 other ptients who, nevertheless, hd mlbsorption of ft or crbohydrte (Tbles 1 nd 2). Notbly, intestinl bsorptive tests my be more sensitive thn serum ntibody tests in monitoring the sttus of intestinl function in dignosed celic sprue. The finding tht more thn hlf of symptomtic celic individuls on gluten-exclusion diet hve ft or crbohydrte mlbsorption suggests strongly tht intestinl dysfunction cused by continuing intestinl injury occurs commonly in the disese, despite lck of typicl clinicl symptoms. This lerts us to identifying incrementl long-term therpy for celic sprue in ddition to the conventionl dietry gluten exclusion, with gol to reduce the reltively high prevlence of osteopeni nd iron-deficiency nemi. Furthermore, the persistence of mlbsorption provides cogent rgument for the regulr clinicl monitoring of celic sprue ptients for intestinl dysfunction. Certinly, those with celic sprue should see n expert t lest once per yer for evlution of intestinl dysfunction; nd n endoscopic biopsy exmintion would be indicted whenever mlbsorption is identified. In those individuls who mounted functionl response to PTC-gluten ingestion, mlbsorption ws voided in the mjority of these gluten-responsive celic ptients when the PTC-gluten ws processed further by PTC-gluten PEP before being consumed (Tbles 3 5). The resons why few ptients were protected indequtely by PEP tretment of gluten re undefined, but the fct tht only the 5-g gluten supplement ws treted with the peptidse should be emphsized. Ingestion of other gluten-contining foods could be the principl reson for the filure to mintin norml bsorptive function in some celic ptients during the ingestion of the PEP-treted gluten. Other possible fctors for the lck of protection is the incomplete PEP specificity for some glidin peptides 5 nd the fct tht other glidin peptide sequences my induce n immune response independent of T-cell ctivtion. 21 Also, pre-existing intestinl dysfunction my reduce brush-border peptidse detoxifying ction fter PEP hs exerted the initil processing. It is encourging to note tht in mny ptients who respond dversely to gluten, pretretment of dietry gluten with PEP results in pprecible bltion of its toxic properties for the smll intestine. Further clinicl
694 PYLE ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 7 studies re needed to ssess the potentil role for n orl peptidse supplement s supportive therpy in ptients with celic sprue. References 1. Sollid LM. Coelic disese: dissecting complex inflmmtory disorder. Nt Rev Immunol 2002;2:647 655. 2. Green PHR, Jbri B. Coelic disese. Lncet 2003;362:383 391. 3. Shn L, Molberg Ø, Prrot I, et l. Structurl bsis for gluten intolernce in celic sprue. Science 2002;297:2275 2279. 4. Husch F, Shn L, Sntigo NA, et l. Intestinl digestive resistnce of immunodominnt glidin peptides. Am J Physiol 2002; 284:G996 G1003. 5. Mrti T, Molberg Ø, Qing L, et l. Prolyl endopeptidse medited destruction of T cell epitopes in whole gluten chemicl nd immunologicl chrcteriztion. J Phrmcol Exp Ther 2005;312: 19 26. 6. Pyle GG, Pso B, Anderson BE, et l. Low-dose gluten chllenge in celic sprue: mlbsorptive nd ntibody responses. Clin Gstroenterol Heptol 2005;3:679 686. 7. Shn L, Mrti T, Sollid LM, et l. Comprtive biochemicl nlysis of three bcteril prolyl endopeptidses: implictions for celic sprue. Biochem J 2004;383:311 318. 8. Portney LG, Wtkins MP. Foundtions of clinicl reserch: pplictions to prctice. Norwlk, CT: Appleton & Lnge, 1993. 9. Koster-Kmphuis L, vn Strten EA, Kors WA, et l. Urinry NOx:cretinine rtios during gluten chllenge in children with celic disese. J Peditr Gstroenterol Nutr 2003;36:372 375. 10. Johnston SD, Smye M, Wtson RP. Intestinl permebility tests in coelic disese. Clin Lb 2001;47:143 150. 11. Crbtree JE, Hetley RV, Juby LD, et l. Serum interleukin-2- receptor in coelic disese: response to tretment nd gluten chllenge. Clin Exp Immunol 1989;77:345 348. 12. Cummins AG, Penttil IA, Lbrooy JT, et l. Recovery of the smll intestine in coelic disese on gluten-free diet: chnges in intestinl permebility, smll bowel morphology nd T-cell ctivity. J Gstroenterol Heptol 1991;6:53 57. 13. Arnott ID, Lo SK. The clinicl utility of wireless cpsule endoscopy. Dig Dis Sci 2004;49:893 901. 14. Fsno A, Ctssi C. Current pproches to dignosis nd tretment of celic disese: n evolving spectrum. Gstroenterology 2001;120:636 651. 15. Wong RCW, Steele RH, Reeves GEM, et l. Antibody nd genetic testing in coelic disese. Pthology 2003;35:285 304. 16. Dieterich W, Lg E, Schopper H, et l. Autontibodies to tissue trnsglutminse s predictors of celic disese. Gstroenterology 1998;115:1317 1321. 17. Vhedi K, Mscrt F, Mry JY, et l. Relibility of ntitrnsglutminse ntibodies s predictors of gluten-free diet complince in dult celic disese. Am J Gstroenterol 2003;98:1079 1087. 18. Cicci C, Cvllro R, dell Vlle N, et l. The use of serum ttg-b ssy in ptients on gluten-free diet s mesure of dietetic complince. Gstroenterology 2002;122:588. 19. Dickey W, Hughes DF, McMilln SA. Disppernce of endomysil ntibodies in treted celic disese does not indicte histologicl recovery. Am J Gstroenterol 2000;95:712 714. 20. Stegn-Guidetti C, Grosso SB, Bruno M, et l. Relibility of immunologic mrkers of celic sprue in the ssessment of mucosl recovery fter gluten withdrwl. J Clin Gstroenterol 1996; 23:101 104. 21. Miuri L, Cicci C, Riccirdelli I, et l. Assocition between innte response to glidin nd ctivtion of pthogenic T cells in coelic disese. Lncet 2003;362:30 37. Address requests for reprints to: Gry M. Gry, MD, Celic Sprue Reserch Foundtion, PO Box 61193, Plo Alto, Cliforni 94304-1193. e-mil: ggry@celicsprue.org; fx: (650) 251-9865. Supported in prt by the Mribeth Evelyn Lynn Fellowship (to T.M.). The uthors thnk the Albert Reserch Council (Edmonton, Cnd) for excellent technicl ssistnce with lrge-scle fermenttion for the preprtion of the prolyl endopeptidse enzyme, nd Professor Byron W. Brown Jr (decesed November 30, 2004) for providing ongoing guidnce on the sttisticl nlyses.