Fig 1.
CONSORT Flow Diagram Depicting Randomization and Intervention Allocation.
CONSORT flow diagram illustrates the process of participant randomization and allocation to intervention groups. The diagram shows the number of subjects assigned to each group, with an indication of those who did not receive the allocated intervention and were subsequently lost to follow-up in both groups. One study participant completed 90 days of participation in the placebo group and remained on Aquamin for the full 180-day period after crossing over due to the closure of the clinical research unit during the COVID-19 pandemic. As a result, this participant was analyzed in the study’s placebo (n = 16) and Aquamin 180-day (n = 13) groups except for proteomics.
Fig 2.
Schematic overview of study design.
Study design is presented in graphical form. Subjects were randomized to either placebo or Aquamin. After 90 days of intervention subjects on placebo crossed over to Aquamin, while subjects on Aquamin remained on Aquamin treatment. There were, thus, three groups: i) Aquamin treatment for 180 days, ii) Placebo for 90 days and iii) Aquamin for 90 days.
Table 1.
Serum chemistry (metabolic panel) values at the baseline, Day-90 and Day-180.
Fig 3.
Chemistry and histology endpoints: Aquamin (180-day treatment group) versus placebo.
A. Serum alkaline phosphatase (ALP): Values (U/L) determined as part of the serum chemistry (metabolic) panel. B. Serum intestine-specific alkaline phosphatase (ALPI): Values (ng/mL) determined by ELISA. C. Serum C-reactive protein (CRP): Values (mg/dL) determined in serum. D. Fecal calprotectin (fCAL): Values (µg/g) determined by ELISA (BÜHLMANN) in stool samples. E. Histological assessment of colon biopsies: Values presented as a simplified Geboes score. All values represent treatment group means and standard deviations. Delta values shown beneath each pair of bars indicate percentage change in the post-treatment value relative to the pretreatment value. Statistical significance between pre- and post-treatment values was determined by paired t-test; asterisk (*) indicates p < 0.05. F. Representative histological appearance of colon tissue from one subject in each treatment group, assessed at Day 0 (pre-intervention) and at the end of treatment (post-intervention). Pre-intervention colon biopsies from both subjects show similar histological features of UC, including crypt architectural distortion and infiltration of pro-inflammatory cells. Post-intervention biopsies from the placebo group continued to display these abnormalities, whereas notable improvement was observed in the subject treated with Aquamin. G. Aquamin versus placebo comparison: Pre- and post-treatment differences for each endpoint are plotted for the two groups to show the divergence between them. H. A composite score was generated by combining the five individual endpoint scores. Statistical significance (p < 0.05) between group (composite) values was determined by unpaired t-test.
Fig 4.
DEXA scan endpoints: Aquamin (180-day treatment group).
A. Femoral neck. BMD, BMC and area values were obtained directly from DEXA scans. Hip strength index was calculated according to Yoshikawa et al [36]. B. Lumbar vertebrae. BMD, BMC and area values were read directly from DEXA scans. All values are treatment group means and standard deviations. Delta values shown beneath each pair of bars indicate percentage change in the post-treatment value relative to the pretreatment value. Statistical significance between pre- and post-treatment values was determined by paired t-test; asterisk (*) indicates p < 0.05. C. DEXA scan inserts: Scans of the Upper femora (left) and lumbar spine (L1-L4, right) were performed at Day 0 (pre-intervention) and Day 180 (post-intervention). The graphs at the bottom of these scans display BMD values for both the femora and spine. On the left, Aquamin improved femoral BMD by 3.4% in a 55-year-old patient, while on the right, lumbar BMD improved by 6% in a 63-year-old participant after 180 days of Aquamin intervention.
Fig 5.
Bone-related serum chemistry endpoints: Aquamin (180-day treatment group) versus placebo.
A. Serum osteocalcin: Values (ng/mL). B. Serum TRAP5b: Values (mIU/mL). C. Serum BALP: Values (ng/mL). All values were determined by ELISA (AssayGenie). Values are treatment group means and standard deviations. Delta values shown beneath each pair of bars indicate percentage change in the post-treatment value relative to the pretreatment value. D. Aquamin versus placebo comparison: Pre- and post-treatment differences for each endpoint are plotted for the two participant groups to show the divergence between the groups. E. A composite score was generated by combining the three individual endpoint scores. Statistical significance between group (composite) values was determined by unpaired t-test; asterisk (*) indicates p < 0.05. [Note: For the composite graph, inverse BALP values are plotted because while increased osteocalcin and TRAP5b are associated with improved bone metabolism, decreased BALP is reflective of improved bone metabolism].
Fig 6.
Proteins of interest identified in colon biopsies by directed search of the proteomic database.
Proteins involved in A. Differentiation; B. Barrier formation; C and D. Inflammation and E. Transport are shown. Data are presented as a heatmap illustrating the pre- and post-intervention changes (up- or downregulation) of individual proteins of interest. Values reflect the Log2 abundance ratio relative to the respective pre-intervention levels for placebo (n = 16) and Aquamin treatment (n = 12) for 180 days (180D). Protein expression is also shown for subjects who crossed over from placebo and received Aquamin for the last 90 days (90D) (n = 16). Mean values for individual proteins were compared to their respective pre-intervention values for statistical differences using the limma package. An asterisk (*) indicates a significant difference from the respective baseline expression at p < 0.05.
Fig 7.
A. Top canonical pathways affected (activated or suppressed) by the proteins shown in Fig 6. Pathways were curated by QIAGEN Ingenuity® Pathway Analysis (IPA) and sorted by p-value. The stacked bar chart displays the percentage of proteins in the database compared to the total number of proteins contributing to each pathway. Red bars depict upregulation, and green bars represent downregulation of proteins in each significantly enriched canonical pathway. The total number of proteins involved in each pathway is listed on the right margin of the graph. The left set of bars illustrates pathways influenced by Aquamin (180-day)-responsive proteins, while the right set shows how the same pathways were affected by the placebo. B. Comparative analysis of canonical pathways. Pathways from the three treatment groups were compared using z-score activation and prioritized with hierarchical clustering, visualized in a heatmap to show trends. Increased z-score activation is represented by orange color, while suppression is depicted by dark blue in the heatmap. Concordance is observed between the Aquamin groups (180- and 90-day).
Fig 8.
A subcellular layout of IPA-generated network analysis.
One of the top networks, illustrated in Fig 8, is associated with gastrointestinal disease, inflammatory response, and organismal injury (Network 1). This network comprises 21 molecules from our dataset with a score of 39. Each network is limited to 35 molecules (IPA default setting) to ensure easy interpretability. The likelihood of these molecules being part of the network is indicated by a p-value calculation. The networks were generated by proteins altered by Aquamin (top) and Placebo (bottom). The color-coded figure legend explains the confidence levels associated with the predicted activity and interaction among these molecules, which are based on their measured expression levels. Notably, the network molecular expression and interactions among these molecules in the Aquamin and placebo groups are completely opposite. The figure legends use color coding to represent the expression and interactions with dark and light colors, respectively, to illustrate the predicted differences. The raw data for the network analysis is provided in S8 Table.