Abstract
We applied constrained ordination numerical ecology methods to data produced with a human intestinal tract-specific phylogenetic microarray (the Aus-HIT Chip) to examine the microbial diversity associated with matched biopsy tissue samples taken from the caecum, transverse colon, sigmoid colon and rectum of 10 healthy patients. Consistent with previous studies, the profiles revealed a marked intersubject variability; however, the numerical ecology methods of analysis allowed the subtraction of the subject effect from the data and revealed, for the first time, evidence of a longitudinal gradient for specific microbes along the colorectum. In particular, probes targeting Streptococcus and Enterococcus spp. produced strongest signals with caecal and transverse colon samples, with a gradual decline through to the rectum. Conversely, the analyses suggest that several members of the Enterobacteriaceae increase in relative abundance towards the rectum. These collective differences were substantiated by the multivariate analysis of quantitative PCR data. We were also able to identify differences in the microarray profiles, especially for the streptococci and Faecalibacterium prausnitzii, on the basis of gender. The results derived by these multivariate analyses are biologically intuitive and suggest that the biogeography of the colonic mucosa can be monitored for changes through cross-sectional and/or inception cohort studies.