- Email: [email protected]
Imprint cytology detects floating Brachyspira in human intestinal spirochetosis
Human Pathology (2010) 41, 249–254
www.elsevier.com/locate/humpath
Original contribution
Imprint cytology detects floating Brachyspira in human intestinal spirochetosis☆,☆☆,★ Sho Ogata MD, PhD a,b,⁎, Masaaki Higashiyama MD a , Yoshikazu Adachi DVM, PhDc , Ichiyo Ohara MD, PhDa , Junichiro Nishiyama MD, PhDa , Yasushi Okusa MD, PhDa , Hiroaki Takeo MD, PhD d , Kimiya Sato MD, PhD d , Kuniaki Nakanishi MD, PhDb , Toshiaki Kawai MD, PhDb a
Division of Medicine, Japan Self Defense Forces Hospital Yokosuka, Yokosuka 237-0071, Japan Department of Pathology and Laboratory Medicine, National Defense Medical College, Tokorozawa 359-8513, Japan c Department of Cell Physiology, School of Agriculture, Ibaraki University, Ibaraki 300-0393, Japan d Department of Pathology, Japan Self Defense Forces Central Hospital, Setagaya, Tokyo 154-8532, Japan b
Received 19 June 2009; revised 22 July 2009; accepted 30 July 2009
Keywords: Brachyspira; Cytology; Genotype; Spirochetosis
Summary Human intestinal spirochetosis is a colorectal infectious disease caused by 2 Brachyspira species. Its diagnosis is established by histology, culture, and polymerase chain reaction, but the value of cytologic examination in routine practice remains unclear. In this study, imprint cytology of biopsy specimens was examined for cytologic features specific to human intestinal spirochetosis. Specimens were obtained from 65 colorectal regions (1–3 regions from each case) in 25 ultrastructurally and/or genetically confirmed human intestinal spirochetosis cases (20 with Brachyspira aalborgi, 3 with B pilosicoli, 2 with both genotypes). In cytologic specimens, spirochetes tended to be floating freely within the mucus and intestinal fluid, whereas the “fringe formation” of spirochetes typically observed in histologic specimens was indistinct in cytologic specimens. Spirochetes were identified in 58 regions (89.2%) and 23 cases (92.0%) by cytology, against in 50 regions (76.9%) and 22 cases (88.0%) by histology (no significant differences). In 6 of 8 regions exhibiting positive cytology and negative histology, B pilosicoli was present within the mucus. Hence, B pilosicoli may tend to float in the mucus. In conclusion, cytologic examination would be useful for the routine identification of human intestinal spirochetosis, especially if B pilosicoli is involved. Further, we suggest the existence of differences in biological behavior between these spirochetes. © 2010 Elsevier Inc. All rights reserved.
☆ The contents of this article were presented at the 5th International Conference of Intestinal Spirochaetal Infections in Animals and Humans on June 8, 2009, in Leon, Spain. ☆☆ This work was supported in part by a grant from the Japan Ministry of Defense (to S. O.). ★ The views expressed in this article are those of the authors and do not reflect the official policies of the Japan Ministry of Defense or the Japan Self Defense Forces. ⁎ Corresponding author. Department of Pathology and Laboratory Medicine, National Defense Medical College, 3-2 Namiki, Tokorozawa, Saitama 359-8513, Japan. E-mail address: [email protected] (S. Ogata).
0046-8177/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2009.07.020
250
S. Ogata et al.
1. Introduction
2.2. Sample preparation in cytology and histology
Human intestinal spirochetosis (HIS) is an infectious disease caused by two Brachyspira species, Brachyspira pilosicoli and/or B aalborgi, and exclusively involving the colorectum [1,2]. In the western world, HIS is considered to have a low prevalence, about 4%, in the general population, and is diagnosed usually by histologic examination of the endoscopic biopsy specimens [3]. However, a definitive diagnosis requires use of polymerase chain reaction method and/or electron microscopy [4] because of the difficulty of identifying spirochetes on the surface epithelium by light microscopy and because the low incidence of the disease means that it is often unrecognized. Although attempts at diagnosis have been made by immunohistochemistry using a polyclonal rabbit antiserum [5] and by the microscopic agglutinin test for serological diagnosis [6], these methods have not yet proved universally reproducible. Therefore, an easy method for detection of spirochetes in routine practice is still required. We investigated the cytologic features of 65 colonic-region samples (from 25 HIS cases with genotypic confirmation) and found that cytologic examination may be useful for the diagnosis and follow-up observation of HIS. Furthermore, our results suggest the existence of biological differences between the bacterial species, and we propose the new idea that HIS can be subdivided according to the floating or attached status of the spirochetes.
In each case, imprint cytology was obtained from 1–3 colonic regions (cecum, transverse colon, and/or sigmoid colon; median region number being 3). Endoscopic biopsy samples were directly imprinted onto glass slides for wetfixed and/or rapidly dried specimens. Wet-fixed (by ethanol) slides for Papanicolaou (Pap) and hematoxylineosin (HE) staining were prepared from a total of 12 regions (5 cases), whereas rapidly dried slides (fixed using methanol) for Giemsa-staining were prepared from all 65 regions (25 cases). Each staining method was performed as described elsewhere [8], with the slight modification of longer immersion with Giemsa solution for Giemsa staining. Just after imprinting for cytology, biopsy specimens were formalin-fixed and paraffin-embedded for routine histology. The cytology and histology specimens were divided into 2 groups based on whether spirochetes were or were not detected (−, spirochetes not found; +, spirochetes found).
2. Materials and methods 2.1. Patient background and diagnostic criteria for HIS Twenty-five patients who underwent endoscopic biopsy procedures during colonoscopy at the Japan Self Defense Forces (JSDF) Hospital Yokosuka (Kanagawa, Japan) from August 2006 to November 2007 were diagnosed with HIS by histology. Patients who had had antibacterial treatment were excluded from the present study. The patients of this hospital are mostly military personnel of the Japan Ministry of Defense and JSDF (and their family members), and display strong male preponderance. The present patients were all men 30 to 53 years old (median age, 40 years). One patient was positive for the antibody for human immunodeficiency virus [7]. Some patients had abdominal pain or diarrheic symptoms, but most underwent endoscopy as part of their annual medical checkup and were asymptomatic. These studies were designed to conform to the Helsinki declaration, and written consent was obtained from all patients. The studies were approved by the institutional ethics committee of JSDF Hospital Yokosuka.
2.3. Bacterial culture and isolation, and polymerase chain reaction method For bacterial culture and isolation, biopsy samples were incubated anaerobically using the BBL GasPak System (Becton, Dickinson and Company, Franklin Lakes, NJ), with the incubation being conducted at 37°C in trypticase soy agar with 5% sheep blood and 400 μg/mL spectinomycin. By repeated subculture using trypticase soy agar with 5% sheep blood, spirochetes were successfully isolated. Total bacterial DNA was extracted from biopsy samples and isolated using InstaGene Matrix (Bio-Rad Laboratories; Hercules, CA). Brachyspira 16S ribosomal DNA for B pilosicoli and B aalborgi were amplified using the corresponding DNA primers, as previously described [9]. polymerase chain reaction (PCR) products for B aalborgi and B pilosicoli 16S ribosomal DNA were detected at 474 and 198 bp, respectively. Cases in which a single genotype was detected in at least 1 region were considered to represent B aalborgi cases or B pilosicoli cases, whereas cases in which both genotypes were detected were considered to represent mixed infection.
2.4. Sample preparation in the ultrastructural study The ultrastructural study was performed, on 10 of the histology-positive cases (see Tables 1 and 2), as described elsewhere [10]. Briefly, cubic biopsy samples were prefixed using glutaraldehyde and postfixed using osmium tetraoxide, serially exchanged with propylene oxide, and embedded in epon resin. Thin sections were stained with uranyl acetate and lead citrate and observed using a transmission electron microscope (H-7500, acceleration
Imprint cytology of intestinal spirochetosis Table 1 results
251
Summary of cases with matching cytology/histology
Case Region of biopsy Genotypes EM No. Cecum Transverse Sigmoid B B colon colon aalborgi pilosicoli Cases diagnosed by both methods 1 + + + – 2 + + + + 3 + + + + 4 + + + + 5 + + + + 6⁎ + + + + 7 + + + + 8 + + + + 9 + + + + 10 + + + + 11 + + + + 12 + + + + 13 + ND + + 14 ND + ND + 15 ND + ND + 16 ND + ND + 17 ND + ND + Cases not diagnosed by either method 18 – – – + 19 – – – +
+ – – – – – – – – – – – – – – – –
ND + + + + + ND ND ND ND ND ND ND + + + +
– –
ND ND
Abbreviations: EM, electron microscopy; ND, not done. ⁎ Case 6 was positive for the serum antibody for human immunodeficiency virus [8].
3. Results 3.1. Histologic aspects of HIS Materials stained with HE revealed a typical histologic feature of HIS, namely hematoxylinophilic comb-like structure (“fringe formation”) on the luminal surface of the colorectal surface epithelium (Fig. 1A). In Giemsa-stained slides, these typical structures were stained purple (Fig. 1B). Furthermore, the spirochetes were reactive with periodic acid–Schiff, but not with the Gram method.
3.2. Cytologic aspects of HIS Epithelial clusters from the intestinal surface and crypt epithelium were seen on the wet-fixed slides (Pap or HE staining) and on the quickly dried slides (Giemsa-staining). “Fringe formations” could be found on the luminal side of the surface epithelium but could not be discriminated from thick microvilli in Pap-, HE-, or Giemsa-stained specimens (Fig. 1C-F). Strikingly, mucus was abundantly attached around the epithelial clusters on the quickly dried slides, and many spirochetes were detected as characteristically spiral microorganisms within the mucus stained by the Giemsa method (Fig. 1E), but fewer were detected in wetfixed specimens.
3.3. Comparison between cytology and histology voltage 80 kV; Hitachi, Tokyo, Japan). Spirochetes were observed as spiral organisms attaching to the colorectal surface epithelium.
2.5. Statistical analysis Statistical analyses were performed by Fisher exact probability test using StatView 5.0 (SAS Institute, Cary, NC), with P b .05 being considered significant.
Table 2 Case No.
Summary of cases with non-matching cytology/histology results Region of biopsy Cecum
20 21 22 23 24 25
For the comparison of imprint cytology with histology in the same biopsy specimens, Giemsa staining was used because of its advantages in detecting HIS (see above). The results are summarized in Tables 1 and 2. Spirochetes were detected in 89.2 % of the 65 regions and in 92.0 % of the 25 cases by cytology, against 76.9% and 88.0%, respectively, by histology. Statistical analysis revealed that regional positivity was not different between cytology and histology (P = .10 by Fisher exact probability test).
Genotypes Transverse colon
Sigmoid colon
Cyto
Histo
Cyto
Histo
Cyto
Histo
+ + + + + ND
+ + + + – ND
+ + + + + +
– – + + + –
+ + + + + ND
– – – – + ND
Abbreviations: Cyto, Cytology; Histo, Histology; EM, electron microscopy; ND, not done.
EM
B aalborgi
B pilosicoli
– – + + + +
+ + + + – –
ND ND + ND ND ND
252
S. Ogata et al.
Fig. 1 Histology (A, B), cytology (C-E), isolated bacteria (F), and ultrastructure (G, H) in HIS. Histology of colonic biopsy specimens revealed hematoxylinophilic thickening of microvilli (“fringe,” reflecting collections of spirochetes; arrowheads) on the luminal surface of the colonic surface epithelial cells [HE staining (A) and Giemsa staining (B)]. Cytologic examination of HIS specimens using wet-fixed preparations (C, D) revealed that typical fringes on the surface epithelium were indistinct in such preparations (C, Pap: original magnification ×40; D, HE: original magnification ×40). In quickly dried specimens, typical fringes on the epithelial clusters were difficult to find (E, Giemsa: original magnification ×40). However, spirochetes were easily observed within the mucus around the epithelial clusters. After culture and isolation, spirochetes were observed by phase contrast microscopy (F: original magnification ×40). Under the electron microscope, typical spiral microorganisms (arrowheads in G) were located perpendicular to the luminal side of the colorectal surface epithelium (G: original magnification ×3500; H: original magnification ×25,000). Scale bars indicate 10 μm in A-F, 5 μm in G, and 1 μm in H.
3.4. Culture, isolation, and PCR for B aalborgi and B pilosicoli Spirochetes were successfully cultured and isolated from every site examined in all 25 cases (Fig. 1F). By PCR, all 25
cases were positive for Brachyspira species (20 for B aalborgi alone, 3 for B pilosicoli alone, and 2 with mixed infections). In 2 cases (cases 18 and 19 in Table 1, both of which involved B aalborgi), spirochetes were not detected either by cytology or by histology.
Imprint cytology of intestinal spirochetosis
3.5. Ultrastructure Typical spiral microorganisms were observed to be arranged perpendicular to the luminal side of the colorectal surface epithelium in all 10 of the cases that were examined by electron microscopy (Fig. 1G-H, Tables 1 and 2). Tissue invasion by spirochetes was not evident.
3.6. Genotypic analysis of cases with non-matching cytology/histology results The results obtained by cytology did not match those obtained by histology in 6 of the 25 cases (24.0%; Table 2), all 6 being cytology-positive and histology-negative in 1 or 2 regions. Concerning the genotypes of these 6 cases, 2 cases were classified as B aalborgi alone, 2 as B pilosicoli alone, and 2 as mixed infections. Our bacterial genotype results indicated that such “result non-matching” cases involved B pilosicoli surprisingly often. To be precise, 4 of the 5 B pilosicoli–positive cases (ie, 80%) were in the result nonmatching group, whereas only 2 of the 20 B. pilosicolinegative cases (ie, 10%) were in that group (P b .01 by Fisher exact probability test).
4. Discussion Although several studies have investigated HIS by histology, electron microscopy, and/or microbiology using culture and molecular methods [3,4,11], there has been no report dealing with the cytology of HIS. In the present study, we found spirochetes within the mucus and intestinal fluid of HIS patients. In these fluids, the spirochetes displayed spiral bacterial morphologies typical of Brachyspira species, a finding that was not revealed by our histologic examination (which showed collections of bacteria as a “fringe formation” on the colorectal surface epithelium). Thus, cytology revealed bacterial features more clearly than histology. Among the diagnostic methods for HIS, bacterial culture of Brachyspira species, as a gold standard method, is difficult and time-consuming (it usually takes 1-4 weeks and may not be successful) [8,11]. In contrast, examination by PCR is a fast tool for the diagnosis of HIS. This method will provide reliable information as to whether spirochetes exist or not. However, it is costly. Cytologic technique using Giemsa stain can overcome these weak points because (i) it is easy, rapid, and involves little cost; (ii) it can highlight the characteristic spirochetal structures (spirally microorganisms with two or more turns), which are usually seen only after performing successful culture and isolation; and (iii) it is able to detect spirochetes within the mucus (which are undetectable by histology), which is important because spirochetes are widely distributed within the mucus of the colorectum. On these grounds, we propose that cytology may be useful and perhaps adequate for the diagnosis and follow-up
253 observation of HIS in routine practice. However, cytology alone, such as histology, is not sufficient for exclusive diagnosis because these morphology-based examinations cannot exclude the possibility of false positives. For example, organisms such as campylobacter might be confused with spirochetes in cytologic examinations. Therefore, confirmation of Brachyspira by another method, such as culture- and PCR-based genotyping, is required before or after cytologic detection. Recently, direct PCR on DNA extracted from the feces has been reported for the diagnosis of HIS [3]. Thus, a combination of cytologic technique and direct fecal PCR may be a useful tool for the purpose of initial identification since it should allow detection of spirochetes within the mucus and intestinal fluid. To judge from the present results, HIS might be classified according to at least 2 states, based on the spirochete biology in human intestines: a floating state and an attached state. In the floating state, spirochetes are floating within the mucus and intestinal fluid layers. In the attached state, in contrast, spirochetes are attached and represent a massive burden on the surface epithelium. In addition to these 2 states, an invasive state (involving tissue invasion by spirochetes) may exist in immunocompromized hosts such as those with acquired immunodeficiency syndrome [12]. It is well known that a spirochete can swim in highly viscous environments with the aid of its periplasmic flagella (spirochetes' organelle of motility). Recently, Nakamura et al [13] suggested that Brachyspira species may have the ability to move into the mucus layer from the intestinal fluid more easily than other bacteria (based on measurements of the ratio of swimming speed to flagellar rotation rate). Thus, the mucus layer may have a special significance for spirochetes. The above observations lead us to speculate that, in HIS, spirochetes may be widely distributed (and proliferate) within the mucus and intestinal fluid of the entire colorectum (floating state) but sometimes attach to the host surface epithelium either focally or diffusely (attached state). Indeed, alternations between these two states may occur, probably reflecting such factors as bacterial attachment capacity, total bacterial burden, host immunity, host microvilli maturation, and/or competition with other bacteria and therapeutic interventions. In the floating state, moreover, there may be some differences in spirochete biology between the mucus layer and the intestinal fluid layers since a very more recent study demonstrated that the mucus layer contains antimicrobial peptides and proteins, and these two layers may possess hostimmunity mechanisms [14]. Hence, cytologic investigations focusing on the mucus layer and intestinal fluid could represent an important avenue for future HIS research. Additional suggestions arise from our analysis of those cases in which cytology and histology results showed a mismatch. Genotypic analysis by PCR raised the possibility of biological differences between B aalborgi and B pilosicoli, although this awaits confirmation. Cases infected by B pilosicoli, either alone or together with B aalborgi, included cases with mismatching cytology/histology results
254 more frequently than those not involving B pilosicoli (ie, cases infected by B aalborgi alone) (4 mismatching cases among 5 cases versus 2 among 20). This might suggest that B pilosicoli had a weaker ability to attach to the human colonic surface epithelium than B aalborgi. In other words, B pilosicoli might be more likely to be in the floating state and B aalborgi in the attached state in HIS. In an experimental animal model, B pilosicoli was found to be present within the crypt mucus more abundantly than on the epithelium [15]. In human colonic biopsies, B aalborgi is frequently detected by PCR using DNA obtained from paraffin-embedded tissue samples, and such samples seldom contain either mucus or intestinal fluid due to the methods of preparation [16,17]. Hence, HIS caused by B pilosicoli may be overlooked by histologic examination (especially by single biopsy examination), and assessments of prevalence based on such tissue-based analysis may underestimate the true prevalence. These inferences from the present study underline the potential importance of genotypic determination of spirochetes on the tissues or within the mucus and intestinal fluid. However, the present study is preliminary for the above hypothesis because of the small number of cases with B pilosicoli. Moreover, in some cases with B pilosicoli in the non-matching cytology/histology group, the histology was negative but the cytology positive for the both the transverse colon and the sigmoid colon, although we detected spirochetes by both methods in the cecum, a more proximal site. Thus, it is possible that this more proximal site was the primary site of colonization by B pilosicoli and that spirochetes were being detected in the mucus at the more distal sites after being shed from the primary site. Actually, we do not know whether Brachyspira in the human intestines proliferate on the epithelium or within the mucus and intestinal fluid. Therefore, our hypothesis must remain untested until more data on HIS is available and more in known about the biology of B pilosicoli. In conclusion, we have demonstrated the potential usefulness of imprint cytology for the detection of HIS, and we propose its application in routine practice. Our data suggest that B pilosicoli may have biologic differences from B aalborgi. Future research should focus on bacterialgenotype determination and on cytology (which can detect spirochetes within the mucus and intestinal fluid layers). This should help to elucidate the true epidemiology, natural history, and pathogenesis of HIS.
Acknowledgments The authors thank Naoko Niihara, Hitoshi Kaga, Akio Kinoshita, and Shinichi Watanabe for excellent technical assistance and useful comments during the cytologic examination and ultrastructural examinations. The authors
S. Ogata et al. also thank RADM Junichi Hatada for his encouragement both before and during this work.
References [1] Rotterdam H. Intestinal spirochetosis. In: Connor DH, Chandler FW, Schwartz DA, et al, editors. Pathology of infectious diseases, Vol. 1. Stamford (Conn): Appleton & Lange; 1997. p. 583-8. [2] Korner M, Gebbers JO. Clinical significance of human intestinal spirochetosis—a morphological approach. Infection 2003;31:341-9. [3] Peruzzi S, Gorrini C, Piccolo G, Calderaro A, Dettori G, Chezzi C. Human intestinal spirochaetosis in Parma: a focus on a selected population during 2002-2005. Acta Biomed 2007;78:128-32. [4] Weisheit B, Bethke B, Stolte M. Human intestinal spirochetosis: analysis of the symptoms of 209 patients. Scand J Gastroenterol 2007; 42:1422-7. [5] Willen R, Wanders A, Engstrand L. Intestinal spirochaetosis. Histopathology 2006;49:658-9. [6] Abe Y, Hirane A, Yoshizawa A, Nakajima H, Adachi Y. The specific antibody to Brachyspira aalborgi in serum obtained from a patient with intestinal spirochetosis. J Vet Med Sci 2006;68:1089-91. [7] Higashiyama M, Ogata S, Adachi Y, et al. Human intestinal spirochetosis accompanied by human immunodeficiency virus infection: a case report. Acta Med Okayama 2009;63:217-21. [8] Raber TL, Buckner L. Cytopathology techniques. In: Mikel UV, editor. Armed Forces Institute of Pathology Advanced Laboratory Methods in Histology and Pathology. Washington, DC.: American Registry of Pathology; 1994. p. 207-38. [9] Tasu C, Nakamura S, Tazawa H, Hara H, Adachi Y. Morphological properties of a human intestinal spirochetosis first isolated from a patient with diarrhea in Japan. Microbiol Immunol 2003;47: 989-96. [10] Deerinck TJ, Martone M, Ellisman MH. Preparative methods for transmission electron microscopy. In: Spector DL, Goldman RD, editors. Basic methods in microscopy. protocols and concepts from cells: a laboratory manual. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press; 2006. p. 307-41. [11] Kraaz W, Pettersson B, Thunberg U, Engstrand L, Fellstrom C. Brachyspira aalborgi infection diagnosed by culture and 16S ribosomal DNA sequencing using human colonic biopsy specimens. J Clin Microbiol 2000;38:2555-60. [12] Guccin JG, Benator DA, Zeller J, Termanini B, Saini N. Intestinal spirochetosis and acquired immunodeficiency syndrome: ultrastructural studies of two cases. Ultrastruct Pathol 1995;19:15-22. [13] Nakamura S, Adachi Y, Goto T, Magariyama Y. Improvement in motion efficiency of the spirochete Brachyspira pilosicoli in viscous environments. Biophys J 2006;90:3019-26. [14] Meyer-Hoffert U, Hornef MW, Henriques-Normark B, et al. Secreted enteric antimicrobial activity localized to the mucus surface layer. Gut 2008;57:764-71. [15] Jensen TK, Moller K, Boye M, Leser TD, Jorsal SE. Scanning electron microscopy and fluorescent in situ hybridization of experimental Brachyspira (Serpulina) pilosicoli infection in growing pigs. Vet Pathol 2000;37:22-32. [16] Mikosza A, La T, Brooke CJ, et al. PCR amplification from fixed tissue indicated frequent involvement of Brachyspira aalborgi in human intestinal spirochetosis. J Clin Microbiol 1999;37:2093-8. [17] Mikosza A, La T, de Boer BW, Hampson DJ. Comparative prevalence of Brachyspira aalborgi and Brachyspira (Serpulina) pilosicoli as etiologic agents of histologically identified intestinal spirochetosis in Australia. J Clin Microbiol 2001;39:347-50.
www.elsevier.com/locate/humpath
Original contribution
Imprint cytology detects floating Brachyspira in human intestinal spirochetosis☆,☆☆,★ Sho Ogata MD, PhD a,b,⁎, Masaaki Higashiyama MD a , Yoshikazu Adachi DVM, PhDc , Ichiyo Ohara MD, PhDa , Junichiro Nishiyama MD, PhDa , Yasushi Okusa MD, PhDa , Hiroaki Takeo MD, PhD d , Kimiya Sato MD, PhD d , Kuniaki Nakanishi MD, PhDb , Toshiaki Kawai MD, PhDb a
Division of Medicine, Japan Self Defense Forces Hospital Yokosuka, Yokosuka 237-0071, Japan Department of Pathology and Laboratory Medicine, National Defense Medical College, Tokorozawa 359-8513, Japan c Department of Cell Physiology, School of Agriculture, Ibaraki University, Ibaraki 300-0393, Japan d Department of Pathology, Japan Self Defense Forces Central Hospital, Setagaya, Tokyo 154-8532, Japan b
Received 19 June 2009; revised 22 July 2009; accepted 30 July 2009
Keywords: Brachyspira; Cytology; Genotype; Spirochetosis
Summary Human intestinal spirochetosis is a colorectal infectious disease caused by 2 Brachyspira species. Its diagnosis is established by histology, culture, and polymerase chain reaction, but the value of cytologic examination in routine practice remains unclear. In this study, imprint cytology of biopsy specimens was examined for cytologic features specific to human intestinal spirochetosis. Specimens were obtained from 65 colorectal regions (1–3 regions from each case) in 25 ultrastructurally and/or genetically confirmed human intestinal spirochetosis cases (20 with Brachyspira aalborgi, 3 with B pilosicoli, 2 with both genotypes). In cytologic specimens, spirochetes tended to be floating freely within the mucus and intestinal fluid, whereas the “fringe formation” of spirochetes typically observed in histologic specimens was indistinct in cytologic specimens. Spirochetes were identified in 58 regions (89.2%) and 23 cases (92.0%) by cytology, against in 50 regions (76.9%) and 22 cases (88.0%) by histology (no significant differences). In 6 of 8 regions exhibiting positive cytology and negative histology, B pilosicoli was present within the mucus. Hence, B pilosicoli may tend to float in the mucus. In conclusion, cytologic examination would be useful for the routine identification of human intestinal spirochetosis, especially if B pilosicoli is involved. Further, we suggest the existence of differences in biological behavior between these spirochetes. © 2010 Elsevier Inc. All rights reserved.
☆ The contents of this article were presented at the 5th International Conference of Intestinal Spirochaetal Infections in Animals and Humans on June 8, 2009, in Leon, Spain. ☆☆ This work was supported in part by a grant from the Japan Ministry of Defense (to S. O.). ★ The views expressed in this article are those of the authors and do not reflect the official policies of the Japan Ministry of Defense or the Japan Self Defense Forces. ⁎ Corresponding author. Department of Pathology and Laboratory Medicine, National Defense Medical College, 3-2 Namiki, Tokorozawa, Saitama 359-8513, Japan. E-mail address: [email protected] (S. Ogata).
0046-8177/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2009.07.020
250
S. Ogata et al.
1. Introduction
2.2. Sample preparation in cytology and histology
Human intestinal spirochetosis (HIS) is an infectious disease caused by two Brachyspira species, Brachyspira pilosicoli and/or B aalborgi, and exclusively involving the colorectum [1,2]. In the western world, HIS is considered to have a low prevalence, about 4%, in the general population, and is diagnosed usually by histologic examination of the endoscopic biopsy specimens [3]. However, a definitive diagnosis requires use of polymerase chain reaction method and/or electron microscopy [4] because of the difficulty of identifying spirochetes on the surface epithelium by light microscopy and because the low incidence of the disease means that it is often unrecognized. Although attempts at diagnosis have been made by immunohistochemistry using a polyclonal rabbit antiserum [5] and by the microscopic agglutinin test for serological diagnosis [6], these methods have not yet proved universally reproducible. Therefore, an easy method for detection of spirochetes in routine practice is still required. We investigated the cytologic features of 65 colonic-region samples (from 25 HIS cases with genotypic confirmation) and found that cytologic examination may be useful for the diagnosis and follow-up observation of HIS. Furthermore, our results suggest the existence of biological differences between the bacterial species, and we propose the new idea that HIS can be subdivided according to the floating or attached status of the spirochetes.
In each case, imprint cytology was obtained from 1–3 colonic regions (cecum, transverse colon, and/or sigmoid colon; median region number being 3). Endoscopic biopsy samples were directly imprinted onto glass slides for wetfixed and/or rapidly dried specimens. Wet-fixed (by ethanol) slides for Papanicolaou (Pap) and hematoxylineosin (HE) staining were prepared from a total of 12 regions (5 cases), whereas rapidly dried slides (fixed using methanol) for Giemsa-staining were prepared from all 65 regions (25 cases). Each staining method was performed as described elsewhere [8], with the slight modification of longer immersion with Giemsa solution for Giemsa staining. Just after imprinting for cytology, biopsy specimens were formalin-fixed and paraffin-embedded for routine histology. The cytology and histology specimens were divided into 2 groups based on whether spirochetes were or were not detected (−, spirochetes not found; +, spirochetes found).
2. Materials and methods 2.1. Patient background and diagnostic criteria for HIS Twenty-five patients who underwent endoscopic biopsy procedures during colonoscopy at the Japan Self Defense Forces (JSDF) Hospital Yokosuka (Kanagawa, Japan) from August 2006 to November 2007 were diagnosed with HIS by histology. Patients who had had antibacterial treatment were excluded from the present study. The patients of this hospital are mostly military personnel of the Japan Ministry of Defense and JSDF (and their family members), and display strong male preponderance. The present patients were all men 30 to 53 years old (median age, 40 years). One patient was positive for the antibody for human immunodeficiency virus [7]. Some patients had abdominal pain or diarrheic symptoms, but most underwent endoscopy as part of their annual medical checkup and were asymptomatic. These studies were designed to conform to the Helsinki declaration, and written consent was obtained from all patients. The studies were approved by the institutional ethics committee of JSDF Hospital Yokosuka.
2.3. Bacterial culture and isolation, and polymerase chain reaction method For bacterial culture and isolation, biopsy samples were incubated anaerobically using the BBL GasPak System (Becton, Dickinson and Company, Franklin Lakes, NJ), with the incubation being conducted at 37°C in trypticase soy agar with 5% sheep blood and 400 μg/mL spectinomycin. By repeated subculture using trypticase soy agar with 5% sheep blood, spirochetes were successfully isolated. Total bacterial DNA was extracted from biopsy samples and isolated using InstaGene Matrix (Bio-Rad Laboratories; Hercules, CA). Brachyspira 16S ribosomal DNA for B pilosicoli and B aalborgi were amplified using the corresponding DNA primers, as previously described [9]. polymerase chain reaction (PCR) products for B aalborgi and B pilosicoli 16S ribosomal DNA were detected at 474 and 198 bp, respectively. Cases in which a single genotype was detected in at least 1 region were considered to represent B aalborgi cases or B pilosicoli cases, whereas cases in which both genotypes were detected were considered to represent mixed infection.
2.4. Sample preparation in the ultrastructural study The ultrastructural study was performed, on 10 of the histology-positive cases (see Tables 1 and 2), as described elsewhere [10]. Briefly, cubic biopsy samples were prefixed using glutaraldehyde and postfixed using osmium tetraoxide, serially exchanged with propylene oxide, and embedded in epon resin. Thin sections were stained with uranyl acetate and lead citrate and observed using a transmission electron microscope (H-7500, acceleration
Imprint cytology of intestinal spirochetosis Table 1 results
251
Summary of cases with matching cytology/histology
Case Region of biopsy Genotypes EM No. Cecum Transverse Sigmoid B B colon colon aalborgi pilosicoli Cases diagnosed by both methods 1 + + + – 2 + + + + 3 + + + + 4 + + + + 5 + + + + 6⁎ + + + + 7 + + + + 8 + + + + 9 + + + + 10 + + + + 11 + + + + 12 + + + + 13 + ND + + 14 ND + ND + 15 ND + ND + 16 ND + ND + 17 ND + ND + Cases not diagnosed by either method 18 – – – + 19 – – – +
+ – – – – – – – – – – – – – – – –
ND + + + + + ND ND ND ND ND ND ND + + + +
– –
ND ND
Abbreviations: EM, electron microscopy; ND, not done. ⁎ Case 6 was positive for the serum antibody for human immunodeficiency virus [8].
3. Results 3.1. Histologic aspects of HIS Materials stained with HE revealed a typical histologic feature of HIS, namely hematoxylinophilic comb-like structure (“fringe formation”) on the luminal surface of the colorectal surface epithelium (Fig. 1A). In Giemsa-stained slides, these typical structures were stained purple (Fig. 1B). Furthermore, the spirochetes were reactive with periodic acid–Schiff, but not with the Gram method.
3.2. Cytologic aspects of HIS Epithelial clusters from the intestinal surface and crypt epithelium were seen on the wet-fixed slides (Pap or HE staining) and on the quickly dried slides (Giemsa-staining). “Fringe formations” could be found on the luminal side of the surface epithelium but could not be discriminated from thick microvilli in Pap-, HE-, or Giemsa-stained specimens (Fig. 1C-F). Strikingly, mucus was abundantly attached around the epithelial clusters on the quickly dried slides, and many spirochetes were detected as characteristically spiral microorganisms within the mucus stained by the Giemsa method (Fig. 1E), but fewer were detected in wetfixed specimens.
3.3. Comparison between cytology and histology voltage 80 kV; Hitachi, Tokyo, Japan). Spirochetes were observed as spiral organisms attaching to the colorectal surface epithelium.
2.5. Statistical analysis Statistical analyses were performed by Fisher exact probability test using StatView 5.0 (SAS Institute, Cary, NC), with P b .05 being considered significant.
Table 2 Case No.
Summary of cases with non-matching cytology/histology results Region of biopsy Cecum
20 21 22 23 24 25
For the comparison of imprint cytology with histology in the same biopsy specimens, Giemsa staining was used because of its advantages in detecting HIS (see above). The results are summarized in Tables 1 and 2. Spirochetes were detected in 89.2 % of the 65 regions and in 92.0 % of the 25 cases by cytology, against 76.9% and 88.0%, respectively, by histology. Statistical analysis revealed that regional positivity was not different between cytology and histology (P = .10 by Fisher exact probability test).
Genotypes Transverse colon
Sigmoid colon
Cyto
Histo
Cyto
Histo
Cyto
Histo
+ + + + + ND
+ + + + – ND
+ + + + + +
– – + + + –
+ + + + + ND
– – – – + ND
Abbreviations: Cyto, Cytology; Histo, Histology; EM, electron microscopy; ND, not done.
EM
B aalborgi
B pilosicoli
– – + + + +
+ + + + – –
ND ND + ND ND ND
252
S. Ogata et al.
Fig. 1 Histology (A, B), cytology (C-E), isolated bacteria (F), and ultrastructure (G, H) in HIS. Histology of colonic biopsy specimens revealed hematoxylinophilic thickening of microvilli (“fringe,” reflecting collections of spirochetes; arrowheads) on the luminal surface of the colonic surface epithelial cells [HE staining (A) and Giemsa staining (B)]. Cytologic examination of HIS specimens using wet-fixed preparations (C, D) revealed that typical fringes on the surface epithelium were indistinct in such preparations (C, Pap: original magnification ×40; D, HE: original magnification ×40). In quickly dried specimens, typical fringes on the epithelial clusters were difficult to find (E, Giemsa: original magnification ×40). However, spirochetes were easily observed within the mucus around the epithelial clusters. After culture and isolation, spirochetes were observed by phase contrast microscopy (F: original magnification ×40). Under the electron microscope, typical spiral microorganisms (arrowheads in G) were located perpendicular to the luminal side of the colorectal surface epithelium (G: original magnification ×3500; H: original magnification ×25,000). Scale bars indicate 10 μm in A-F, 5 μm in G, and 1 μm in H.
3.4. Culture, isolation, and PCR for B aalborgi and B pilosicoli Spirochetes were successfully cultured and isolated from every site examined in all 25 cases (Fig. 1F). By PCR, all 25
cases were positive for Brachyspira species (20 for B aalborgi alone, 3 for B pilosicoli alone, and 2 with mixed infections). In 2 cases (cases 18 and 19 in Table 1, both of which involved B aalborgi), spirochetes were not detected either by cytology or by histology.
Imprint cytology of intestinal spirochetosis
3.5. Ultrastructure Typical spiral microorganisms were observed to be arranged perpendicular to the luminal side of the colorectal surface epithelium in all 10 of the cases that were examined by electron microscopy (Fig. 1G-H, Tables 1 and 2). Tissue invasion by spirochetes was not evident.
3.6. Genotypic analysis of cases with non-matching cytology/histology results The results obtained by cytology did not match those obtained by histology in 6 of the 25 cases (24.0%; Table 2), all 6 being cytology-positive and histology-negative in 1 or 2 regions. Concerning the genotypes of these 6 cases, 2 cases were classified as B aalborgi alone, 2 as B pilosicoli alone, and 2 as mixed infections. Our bacterial genotype results indicated that such “result non-matching” cases involved B pilosicoli surprisingly often. To be precise, 4 of the 5 B pilosicoli–positive cases (ie, 80%) were in the result nonmatching group, whereas only 2 of the 20 B. pilosicolinegative cases (ie, 10%) were in that group (P b .01 by Fisher exact probability test).
4. Discussion Although several studies have investigated HIS by histology, electron microscopy, and/or microbiology using culture and molecular methods [3,4,11], there has been no report dealing with the cytology of HIS. In the present study, we found spirochetes within the mucus and intestinal fluid of HIS patients. In these fluids, the spirochetes displayed spiral bacterial morphologies typical of Brachyspira species, a finding that was not revealed by our histologic examination (which showed collections of bacteria as a “fringe formation” on the colorectal surface epithelium). Thus, cytology revealed bacterial features more clearly than histology. Among the diagnostic methods for HIS, bacterial culture of Brachyspira species, as a gold standard method, is difficult and time-consuming (it usually takes 1-4 weeks and may not be successful) [8,11]. In contrast, examination by PCR is a fast tool for the diagnosis of HIS. This method will provide reliable information as to whether spirochetes exist or not. However, it is costly. Cytologic technique using Giemsa stain can overcome these weak points because (i) it is easy, rapid, and involves little cost; (ii) it can highlight the characteristic spirochetal structures (spirally microorganisms with two or more turns), which are usually seen only after performing successful culture and isolation; and (iii) it is able to detect spirochetes within the mucus (which are undetectable by histology), which is important because spirochetes are widely distributed within the mucus of the colorectum. On these grounds, we propose that cytology may be useful and perhaps adequate for the diagnosis and follow-up
253 observation of HIS in routine practice. However, cytology alone, such as histology, is not sufficient for exclusive diagnosis because these morphology-based examinations cannot exclude the possibility of false positives. For example, organisms such as campylobacter might be confused with spirochetes in cytologic examinations. Therefore, confirmation of Brachyspira by another method, such as culture- and PCR-based genotyping, is required before or after cytologic detection. Recently, direct PCR on DNA extracted from the feces has been reported for the diagnosis of HIS [3]. Thus, a combination of cytologic technique and direct fecal PCR may be a useful tool for the purpose of initial identification since it should allow detection of spirochetes within the mucus and intestinal fluid. To judge from the present results, HIS might be classified according to at least 2 states, based on the spirochete biology in human intestines: a floating state and an attached state. In the floating state, spirochetes are floating within the mucus and intestinal fluid layers. In the attached state, in contrast, spirochetes are attached and represent a massive burden on the surface epithelium. In addition to these 2 states, an invasive state (involving tissue invasion by spirochetes) may exist in immunocompromized hosts such as those with acquired immunodeficiency syndrome [12]. It is well known that a spirochete can swim in highly viscous environments with the aid of its periplasmic flagella (spirochetes' organelle of motility). Recently, Nakamura et al [13] suggested that Brachyspira species may have the ability to move into the mucus layer from the intestinal fluid more easily than other bacteria (based on measurements of the ratio of swimming speed to flagellar rotation rate). Thus, the mucus layer may have a special significance for spirochetes. The above observations lead us to speculate that, in HIS, spirochetes may be widely distributed (and proliferate) within the mucus and intestinal fluid of the entire colorectum (floating state) but sometimes attach to the host surface epithelium either focally or diffusely (attached state). Indeed, alternations between these two states may occur, probably reflecting such factors as bacterial attachment capacity, total bacterial burden, host immunity, host microvilli maturation, and/or competition with other bacteria and therapeutic interventions. In the floating state, moreover, there may be some differences in spirochete biology between the mucus layer and the intestinal fluid layers since a very more recent study demonstrated that the mucus layer contains antimicrobial peptides and proteins, and these two layers may possess hostimmunity mechanisms [14]. Hence, cytologic investigations focusing on the mucus layer and intestinal fluid could represent an important avenue for future HIS research. Additional suggestions arise from our analysis of those cases in which cytology and histology results showed a mismatch. Genotypic analysis by PCR raised the possibility of biological differences between B aalborgi and B pilosicoli, although this awaits confirmation. Cases infected by B pilosicoli, either alone or together with B aalborgi, included cases with mismatching cytology/histology results
254 more frequently than those not involving B pilosicoli (ie, cases infected by B aalborgi alone) (4 mismatching cases among 5 cases versus 2 among 20). This might suggest that B pilosicoli had a weaker ability to attach to the human colonic surface epithelium than B aalborgi. In other words, B pilosicoli might be more likely to be in the floating state and B aalborgi in the attached state in HIS. In an experimental animal model, B pilosicoli was found to be present within the crypt mucus more abundantly than on the epithelium [15]. In human colonic biopsies, B aalborgi is frequently detected by PCR using DNA obtained from paraffin-embedded tissue samples, and such samples seldom contain either mucus or intestinal fluid due to the methods of preparation [16,17]. Hence, HIS caused by B pilosicoli may be overlooked by histologic examination (especially by single biopsy examination), and assessments of prevalence based on such tissue-based analysis may underestimate the true prevalence. These inferences from the present study underline the potential importance of genotypic determination of spirochetes on the tissues or within the mucus and intestinal fluid. However, the present study is preliminary for the above hypothesis because of the small number of cases with B pilosicoli. Moreover, in some cases with B pilosicoli in the non-matching cytology/histology group, the histology was negative but the cytology positive for the both the transverse colon and the sigmoid colon, although we detected spirochetes by both methods in the cecum, a more proximal site. Thus, it is possible that this more proximal site was the primary site of colonization by B pilosicoli and that spirochetes were being detected in the mucus at the more distal sites after being shed from the primary site. Actually, we do not know whether Brachyspira in the human intestines proliferate on the epithelium or within the mucus and intestinal fluid. Therefore, our hypothesis must remain untested until more data on HIS is available and more in known about the biology of B pilosicoli. In conclusion, we have demonstrated the potential usefulness of imprint cytology for the detection of HIS, and we propose its application in routine practice. Our data suggest that B pilosicoli may have biologic differences from B aalborgi. Future research should focus on bacterialgenotype determination and on cytology (which can detect spirochetes within the mucus and intestinal fluid layers). This should help to elucidate the true epidemiology, natural history, and pathogenesis of HIS.
Acknowledgments The authors thank Naoko Niihara, Hitoshi Kaga, Akio Kinoshita, and Shinichi Watanabe for excellent technical assistance and useful comments during the cytologic examination and ultrastructural examinations. The authors
S. Ogata et al. also thank RADM Junichi Hatada for his encouragement both before and during this work.
References [1] Rotterdam H. Intestinal spirochetosis. In: Connor DH, Chandler FW, Schwartz DA, et al, editors. Pathology of infectious diseases, Vol. 1. Stamford (Conn): Appleton & Lange; 1997. p. 583-8. [2] Korner M, Gebbers JO. Clinical significance of human intestinal spirochetosis—a morphological approach. Infection 2003;31:341-9. [3] Peruzzi S, Gorrini C, Piccolo G, Calderaro A, Dettori G, Chezzi C. Human intestinal spirochaetosis in Parma: a focus on a selected population during 2002-2005. Acta Biomed 2007;78:128-32. [4] Weisheit B, Bethke B, Stolte M. Human intestinal spirochetosis: analysis of the symptoms of 209 patients. Scand J Gastroenterol 2007; 42:1422-7. [5] Willen R, Wanders A, Engstrand L. Intestinal spirochaetosis. Histopathology 2006;49:658-9. [6] Abe Y, Hirane A, Yoshizawa A, Nakajima H, Adachi Y. The specific antibody to Brachyspira aalborgi in serum obtained from a patient with intestinal spirochetosis. J Vet Med Sci 2006;68:1089-91. [7] Higashiyama M, Ogata S, Adachi Y, et al. Human intestinal spirochetosis accompanied by human immunodeficiency virus infection: a case report. Acta Med Okayama 2009;63:217-21. [8] Raber TL, Buckner L. Cytopathology techniques. In: Mikel UV, editor. Armed Forces Institute of Pathology Advanced Laboratory Methods in Histology and Pathology. Washington, DC.: American Registry of Pathology; 1994. p. 207-38. [9] Tasu C, Nakamura S, Tazawa H, Hara H, Adachi Y. Morphological properties of a human intestinal spirochetosis first isolated from a patient with diarrhea in Japan. Microbiol Immunol 2003;47: 989-96. [10] Deerinck TJ, Martone M, Ellisman MH. Preparative methods for transmission electron microscopy. In: Spector DL, Goldman RD, editors. Basic methods in microscopy. protocols and concepts from cells: a laboratory manual. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press; 2006. p. 307-41. [11] Kraaz W, Pettersson B, Thunberg U, Engstrand L, Fellstrom C. Brachyspira aalborgi infection diagnosed by culture and 16S ribosomal DNA sequencing using human colonic biopsy specimens. J Clin Microbiol 2000;38:2555-60. [12] Guccin JG, Benator DA, Zeller J, Termanini B, Saini N. Intestinal spirochetosis and acquired immunodeficiency syndrome: ultrastructural studies of two cases. Ultrastruct Pathol 1995;19:15-22. [13] Nakamura S, Adachi Y, Goto T, Magariyama Y. Improvement in motion efficiency of the spirochete Brachyspira pilosicoli in viscous environments. Biophys J 2006;90:3019-26. [14] Meyer-Hoffert U, Hornef MW, Henriques-Normark B, et al. Secreted enteric antimicrobial activity localized to the mucus surface layer. Gut 2008;57:764-71. [15] Jensen TK, Moller K, Boye M, Leser TD, Jorsal SE. Scanning electron microscopy and fluorescent in situ hybridization of experimental Brachyspira (Serpulina) pilosicoli infection in growing pigs. Vet Pathol 2000;37:22-32. [16] Mikosza A, La T, Brooke CJ, et al. PCR amplification from fixed tissue indicated frequent involvement of Brachyspira aalborgi in human intestinal spirochetosis. J Clin Microbiol 1999;37:2093-8. [17] Mikosza A, La T, de Boer BW, Hampson DJ. Comparative prevalence of Brachyspira aalborgi and Brachyspira (Serpulina) pilosicoli as etiologic agents of histologically identified intestinal spirochetosis in Australia. J Clin Microbiol 2001;39:347-50.