Fig. S1. Relative proportions of Fusobacterium nucleatum. Data are mean values from positive samples arranged by group and origin (i.e. B = Bhulbule, M = Manang). Shown is the increase (left) or decrease (right) of F. nucleatum upon travelling from Bhulbule to Manang. Percentage values were obtained by dividing rRNA gene copy numbers of F. nucleatum by the rRNA gene copy number of the entire bacterial load. Bars indicate standard error. Further information regarding methodological approach is provided in Supporting information S1.

Fig. S2. Assignment of T-RFs to major bacterial genera. The assignment was made according to the MiCa analysis (‘Virtual Digest’) using the human oral 16S rRNA gene database HOMD containing approximately 600 validated microbial taxa. Shown are observed T-RFs (bp given on top of the columns) that were present in at least 10% of samples. The specificity of the observed T-RFs is categorized in grey scales, with light grey indicating one, medium grey two, dark grey three and black more than four genera to which a given T-RF can potentially be assigned to. Note that in a few cases, rare genera (i.e. those considered to be transient or with less than 1% of prevalence according to Dewhirst et al., 2010) were not included.

Fig. S3. Summarized T-RFLP profiles. Depicted is a condensed excel file in alignment according to the Genotyper software. Each column represents a specific T-RF, while each row represents a distinct individual sample. Grey areas indicate the presence of a T-RF with reduction in relative proportion at the end of the trip. Black areas indicate the presence of a T-RF with elevation of relative proportion at the end of the trip. The horizontal arrows indicate the direction of increasing fragment size of the T-RFs. (A) T-RFs from 40 to 219 bp are shown. (B) T-RFs from 220 bp to 467 bp are shown. Vertical arrows highlight those T-RFs with distinct differences either between groups I and II or between start and end of the trip. For further information, see main text and Supporting information S4.

Fig. S4. AMMI analysis of bacterial T-RFLP data sets. T-RFLP profiles were analysed separately for the two sampled locations (Bhulbule and Manang) but together for groups I and II (A and B) or separately for groups I and II but together for the two sampled location (C and D). Each quadrate represents one T-RFLP profile with the suffix ‘E + number’ referring to one sample. Red dots indicate samples from Group II in (A) and (B), yellow dots indicate samples from Manang in (C) and (D). Variation of bacterial communities from interaction effects (IE) was similarly high in samples collected in Bhulbule or Manang (58.2% versus 60.4%, A + B). Conversely, variation from IE was different between Group I and Group II (53.3% versus 61.8%, C + D), indicating that the bacterial communities in Group I were more homogenous than in Group II.

Fig. S5. AMMI analysis of bacterial T-RFLP data sets. T-RFLP profiles were analysed separately for both different sampling locations and for the two different groups. Each quadrate represents one T-RFLP profile with the suffix ‘E + number’ referring to one individual. T-RFLP profiles compared within Group I were more similar to each other than the T-RFLP profiles within Group II at begin of the trip (IE of 49.6% versus 63.0%, A + B). Accordingly, the first two dimensions (IPCA1 and IPCA2) captured a higher range of the overall variation of T-RFLP profiles in Group I versus Group II (i.e. up to 80% in A versus only 58% in B). After the trip, variations from IE were almost equalized between the groups (59.2% versus 60.9%, C + D).

Fig. S6. AMMI analysis of bacterial T-RFLP data sets from eight individuals. Each quadrate represents one T-RFLP profile with the suffix ‘E + number’ referring to one sample. Circles comprise the two T-RFLP-profiles per individual, with the arrow directing from first sample taken at Bhulbule to the second sample taken at Manang. Closed circles: individuals from Group II; dashed circles: individuals from Group I.

Supporting information S1. Detailed description of material and methods.

Supporting information S2. Comparison of clinical data between Group I and Group II.

Supporting information S3. Number of theoretical cutting sites of different restriction enzymes.

Supporting information S4. Provisional assignment of selected T-RFs to bacterial taxa based on the MiCa analysis of the Human oral microbiome databases (HOMD).

Supporting information S5. Results of AMMI analysis based on an alternative noise filtering algorithm.

emi4340_sm_FigureS1.tif269KSupporting info item
emi4340_sm_FigureS2.tif884KSupporting info item
emi4340_sm_FigureS3.tif552KSupporting info item
emi4340_sm_FigureS4.tif1721KSupporting info item
emi4340_sm_FigureS5.tif1251KSupporting info item
emi4340_sm_FigureS6.tif1409KSupporting info item
emi4340_sm_SupportingS1.doc87KSupporting info item
emi4340_sm_SupportingS2.doc34KSupporting info item
emi4340_sm_SupportingS3.doc31KSupporting info item
emi4340_sm_SupportingS4.doc70KSupporting info item
emi4340_sm_SupportingS5.doc127KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.