Linking peri-implantitis to microbiome changes in affected implants, healthy implants, and saliva: a cross-sectional pilot study

Abstract

Introduction: The rising use of dental implants is accompanied by an expected increase in peri-implant diseases, particularly peri-implantitis (PI), which poses a significant threat to implant success and necessitates a thorough understanding of its pathogenesis for effective management.

Methods: To gain deeper insights into the role and impact of the peri-implant microbiome in the pathogenesis and progression of PI, we analyzed 100 samples of saliva and subgingival biofilm from 40 participants with healthy implants (HI group) or with co-occurrence of diagnosed PI-affected implants and healthy implants (PI group) using shotgun metagenomic sequencing. We identified the most discriminative species distinguishing healthy from diseased study groups through log ratios and differential ranking analyses.

Results and discussion: Mogibacterium timidum, Schaalia cardiffensis, Parvimonas micra, Filifactor alocis, Porphyromonas endodontalis, Porphyromonas gingivalis and Olsenella uli were associated with the subgingival peri-implant biofilm. In contrast, Neisseria sp oral taxon 014, Haemophilus parainfluenzae, Actinomyces naeslundii, Rothia mucilaginosa and Rothia aeria were more prevalent in the healthy peri-implant biofilm. Functional pathways such as arginine and polyamine biosynthesis, including putrescine and citrulline biosynthesis, showed stronger correlations with PI-affected implants. In contrast, peri-implant health was characterized by the predominance of pathways involved in purine and pyrimidine deoxyribonucleotide de novo biosynthesis, glucose and glucose-1-phosphate degradation, and tetrapyrrole biosynthesis. Our findings reveal that healthy implants in PI-free oral cavities differ significantly in microbial composition and functional pathways compared to healthy implants co-occurring with PI-affected implants, which more closely resemble PI-associated profiles. This pattern extended to salivary samples, where microbial and functional biomarkers follow similar trends.

Keywords: compositional change; differential rankings; functional pathways; peri-implant microbiome; peri-implantitis; saliva microbiome; shotgun metagenomic sequencing.

Figure 1

(A) Relative abundances (>1%) of the bacterial composition at the species level in each study group. (B) Hill diversity indices (Species richness, Shannon and Inverted Simpson) of the bacterial communities of the five study groups. *p < 0.05, **p < 0.01, ***p < 0.001, based on the Kruskal-Wallis rank.

Figure 2

Significant taxonomic differences in the oral metagenome of saliva from healthy and PI-affected patients (HI_Sa vs PI_Sa). (A) Bacterial differential ranks of the 23 out of 231 (9.96%) species more (identified as numerator) and less (identified as denominator) associated with PI_Sa using the group HI_Sa as reference, as estimated from multinomial regression by Songbird. (B) Log ratio plots of the 23 out of 231 species across HI_Sa and PI_Sa groups. (C) Log ratio plots of specific combinations of bacterial species across HI_Sa and PI_Sa groups. Statistical significance based on a Student’s t-test (*p < 0.05; **p < 0.01).

Figure 3

Differential ranks of the 10% species changing the most relative to each other in the two compared groups of subgingival biofilms, as estimated from multinomial regression by Songbird. (A) 14 out of 145 species presenting very different ranks in PI_HIS vs PI_PIS. (B) 15 out of 159 species presenting very different ranks in HI_HIS vs PI_PIS. (C) 13 out of 137 species presenting very different ranks in HI_HIS vs PI_HIS.

Figure 4

Log ratios of the top and bottom 10% species changing the most across the two groups in comparison. Statistical significance based on a Student’s t-test (**p < 0.01).

Figure 5

Log ratio plots of specific combinations of bacterial species across HI_HIS and PI_PIS groups (A), PI_HIS and PI_PIS groups (B), and HI_HIS and PI_HIS (C). Statistical significance based on a Student’s t-test (*p < 0.05; **p < 0.01).

Figure 6

Log ratio plots of the top and bottom 10% functional pathways selected across HI_Sa and PI_Sa (A), and log ratio plots of specific combinations of functional pathways across HI_Sa and PI_Sa (B). [a): log (ARG+POLYAMINE-SYN: superpathway of arginine and polyamine biosynthesis + PWY-5005: biotin biosynthesis II + PWY-822: fructan biosynthesis/PWY-1269: CMP-3-deoxy-D-manno-octulosonate biosynthesis + PANTO-PWY: phosphopantothenate biosynthesis I + PWY-5840: superpathway of menaquinol-7 biosynthesis); b): log (PWY-6305: superpathway of putrescine biosynthesis + PWY-7254: TCA cycle VII (acetate-producers) + PWY-822: fructan biosynthesis/PWY-1269: CMP-3-deoxy-D-manno-octulosonate biosynthesis + PANTO-PWY: phosphopantothenate biosynthesis I + PWY-5840: superpathway of menaquinol-7 biosynthesis)]. Statistical significance based on a Student’s t-test (*p < 0.05).

Figure 7

Log ratio plots of the top and bottom 10% functional pathways selected across PI_HIS and PI_PIS (A), and log ratio plots of specific combinations of functional pathways across PI_HIS and PI_PIS (B). [a): log (BIOTIN-BIOSYNTHESIS-PWY: biotin biosynthesis I + CITRULBIO-PWY: L-citrulline biosynthesis + ARG+POLYAMINE-SYN: superpathway of arginine and polyamine biosynthesis/P124-PWY: Bifidobacterium shunt + PWY-7013: (S)-propane-1,2-diol degradation + PWY-7210: pyrimidine deoxyribonucleotides biosynthesis from CTP); b): log (BIOTIN-BIOSYNTHESIS-PWY: biotin biosynthesis I + PWY-6305: superpathway of putrescine biosynthesis + CITRULBIO-PWY: L-citrulline biosynthesis/P124-PWY: Bifidobacterium shunt + PWY-7013: (S)-propane-1,2-diol degradation + PWY-7210: pyrimidine deoxyribonucleotides biosynthesis from CTP); c): log (ARG+POLYAMINE-SYN: superpathway of arginine and polyamine biosynthesis + PWY-6305: superpathway of putrescine biosynthesis + CITRULBIO-PWY: L-citrulline biosynthesis/P124-PWY: Bifidobacterium shunt + P185-PWY: formaldehyde assimilation III (dihydroxyacetone cycle) + PWY-7210: pyrimidine deoxyribonucleotides biosynthesis from CTP]. Statistical significance based on a Student’s t-test (*p < 0.05).

Figure 8

Log ratio plots of the top and bottom 10% functional pathways selected across HI_HIS and PI_PIS (A), and log ratio plots of specific combinations of functional pathways across HI_HIS and PI_PIS (B). [a): log (ARGININE-SYN4-PWY: L-ornithine biosynthesis II + CITRULBIO-PWY: L-citrulline biosynthesis + PWY0-1297: superpathway of purine deoxyribonucleosides degradation/GLUCOSE1PMETAB-PWY: glucose and glucose-1-phosphate degradation + PWY-7013: (S)-propane-1,2-diol degradation + PWY-7883: anhydromuropeptides recycling II); b): log (P164-PWY: purine nucleobases degradation I (anaerobic) + ARG+POLYAMINE-SYN: superpathway of arginine and polyamine biosynthesis + PWY-5838: superpathway of menaquinol-8 biosynthesis I/PWY-5189: tetrapyrrole biosynthesis II (from glycine) + GLUCOSE1PMETAB-PWY: glucose and glucose-1-phosphate degradation + DENOVOPURINE2-PWY: superpathway of purine nucleotides de novo biosynthesis II); c): log (P164-PWY: purine nucleobases degradation I (anaerobic) + ARG+POLYAMINE-SYN: superpathway of arginine and polyamine biosynthesis + CITRULBIO-PWY: L-citrulline biosynthesis/PWY-5189: tetrapyrrole biosynthesis II (from glycine) + GLUCOSE1PMETAB-PWY: glucose and glucose-1-phosphate degradation + DENOVOPURINE2-PWY: superpathway of purine nucleotides de novo biosynthesis II]. Statistical significance based on a Student’s t-test (*p < 0.05; **p < 0.01).

Figure 9

Log ratio plots of the top and bottom 10% functional pathways selected across HI_HIS and PI_HIS (A). Log ratio plots of a specific combination of functional pathways across HI_HIS and PI_HIS (B): log [PWY-5838: superpathway of menaquinol-8 biosynthesis I + PWY0-1297: superpathway of purine deoxyribonucleosides degradation + P124-PWY: Bifidobacterium shunt/PWY-5918: superpathway of heme b biosynthesis from glutamate + PWY-8131: 5’-deoxyadenosine degradation II + PWY-5189: tetrapyrrole biosynthesis II (from glycine)]. Statistical significance based on a Student’s t-test (*p < 0.05).