Microbial Bioactives

1. Introduction

In the first decades of the 21st century, the rise of antimicrobial resistance (AMR) has transformed once-treatable infections into formidable clinical challenges. Bacterial pathogens, especially notorious groups like the ESKAPE organisms, continually evolve mechanisms that reduce the efficacy of conventional antibiotics, pushing global health systems toward a precipice (World Health Organization [WHO], 2014). Our reliance on single-target synthetic antibiotics has created an evolutionary arms race; pathogens adapt faster than new drugs can be developed, leading to a critically thin pipeline of novel agents (Newman & Cragg, 2016). In response, scientific inquiry has increasingly turned toward nature’s chemical diversity—exploring plant secondary metabolites and other natural products—as a complementary strategy to conventional antimicrobials (Cowan, 1999; Savoia, 2012; Vaou et al., 2022).

Unlike synthetic agents engineered to attack a single molecular target, plant-derived compounds emerge from millions of years of evolutionary refinement. These molecules exist in complex mixtures that plants themselves use to defend against microbes, herbivores, and environmental stressors (Ganora, 2009). This intrinsic complexity means that therapeutic effects often arise not from a single constituent, but from dynamic interactions among many bioactives—interactions that may be synergistic, additive, or antagonistic (Rather, Bhat, & Qurishi, 2013; Wagner & Ulrich-Merzenich, 2009). Among these, synergistic combinations—where the combined effect of compounds exceeds the sum of their individual actions—are of particular interest because they offer the potential to enhance antimicrobial potency while reducing toxicity and resistance pressure (Hemaiswarya, Kruthiventi, & Doble, 2008; Vaou et al., 2022).

Plant secondary metabolites such as polyphenols, flavonoids, alkaloids, terpenes, and essential oils have demonstrated broad mechanisms of microbial inhibition, including membrane disruption, enzyme inhibition, interference with nucleic acid synthesis, and efflux pump modulation (Cowan, 1999; Daglia, 2012; Górniak, Bartoszewski, & Króliczewski, 2019). Yet individual compounds often fall short when tested alone, particularly against complex, multidrug-resistant (MDR) isolates. It is the strategic combination of these compounds—guided by systematic evaluation and meta-analytic synthesis—that reveals their true potential (Rather et al., 2013; Vaou et al., 2022). For example, flavonoids such as baicalein and 7-hydroxyflavone have shown potentiating effects when paired with conventional antibiotics, suggesting that plant molecules can modify pathogen susceptibility to synthetic drugs (Jang et al., 2014; Tang, Wennerberg, & Aittokallio, 2015).

To understand why combinations of natural products matter, we must first appreciate the diversity of interaction types. Synergy arises when compounds enhance each other’s antimicrobial action in a way that the combined effect is greater than expected. Mechanistically, this can occur through pharmacodynamic synergy—where different molecular targets are hit simultaneously—or pharmacokinetic synergy, where one compound enhances the absorption or distribution of another (Hemaiswarya et al., 2008; Wagner & Ulrich-Merzenich, 2009). Additive interactions reflect pure summation of independent effects, while antagonistic interactions reduce the overall efficacy and must be identified to avoid clinical harm (Berenbaum, 1989; Langeveld, Veldhuizen, & Burt, 2014). Quantitative tools such as the Fractional Inhibitory Concentration (FIC) index and isobologram analysis have become standard in systematic studies evaluating these interactions (Vaou et al., 2022).

A growing body of literature demonstrates that paired or multi-component phytochemical therapies can limit the emergence of resistance. Classic examples include berberine combined with pump-inhibiting flavonols to overcome efflux-mediated resistance in Staphylococcus aureus (Stermitz, Lorenz, Tawara, Zenewicz, & Lewis, 2000), and peppermint oil components that enhance antibiotic activity against respiratory pathogens (Bassolé & Lamien-Meda, 2010). Similarly, cannabinoids such as cannabidiol (CBD) have been identified as potent antibiotic adjuvants, reversing resistance in Gram-positive bacteria when combined with traditional drugs like bacitracin (Blaskovich et al., 2021; Karas et al., 2020). These findings reflect the principle that complex microbial defenses can be dismantled more effectively by coordinated multi-target attacks than by single-agent therapies.

The systematic integration of these data highlights several strengths of plant-derived combinations. First, synergistic mixtures often allow dose reduction of both plant extracts and conventional drugs, which can reduce host toxicity and side effects (Vaou et al., 2022). Second, the likelihood of resistance emergence declines when microbes must simultaneously adapt to multiple distinct mechanisms of action (Hemaiswarya et al., 2008; Wagner & Ulrich-Merzenich, 2009). Third, the diversity of plant chemistry provides a vast reservoir of unexplored bioactivity, making natural sources a fertile ground for novel adjuvants that can restore obsolete antibiotics to clinical utility (Newman & Cragg, 2016; Savoia, 2012).

Despite this promise, translating plant combinations into clinical practice requires rigorous systematic evaluation. Unlike single-compound pharmaceuticals, botanical extracts can vary widely based on species, geography, harvest time, and extraction method, all of which influence chemical composition (Eloff, 2004; van Vuuren & Viljoen, 2011). Thus, meta-analytic approaches that synthesize data across multiple studies provide essential insight into reproducibility and effect consistency. For example, meta-analysis of minimum inhibitory concentration (MIC) values across trials reveals patterns of potency and variability that inform both mechanistic understanding and prioritization for further development (Vaou et al., 2022).

The adoption of advanced analytical methods such as metabolomics and biochemometrics further enhances systematic discovery. These tools allow researchers to correlate specific compounds or clusters of metabolites with antimicrobial activity, effectively deconvoluting the complexity inherent in plant extracts (Caesar, Kellogg, Kvalheim, & Rønsted, 2018; Nakabayashi & Saito, 2013). When integrated with network-based analyses and computational synergy prediction models, researchers can identify promising combinations before extensive in vitro or in vivo validation (Tang et al., 2015).

Model organisms such as Caenorhabditis elegans have also become integral to the translational pipeline, enabling in vivo testing of phytochemical combinations for both efficacy and toxicity within a living organism (Moy et al., 2006). These systems bridge the gap between isolated in vitro assays and complex vertebrate models, offering an ethically and economically tractable step in the path toward clinical relevance (Moy et al., 2006).

While plant combinations show great promise, they are not without limitations. Antagonistic interactions, though less desirable, must be systematically excluded to avoid counterproductive outcomes (Berenbaum, 1989). Additionally, the majority of current research remains preclinical; few combinations have advanced to human trials, underscoring the need for continued, rigorously designed translational studies (van Vuuren & Viljoen, 2011).

In summary, the systematic study of plant-derived antimicrobial combinations reveals a rich and underutilized resource in the fight against AMR. By leveraging synergy, additive interactions, and advanced analytical techniques, natural products offer a multi-targeted arsenal against resistant pathogens. When coupled with meta-analytic synthesis, these approaches can guide the development of new therapeutic strategies that enhance efficacy, reduce resistance pressure, and broaden the utility of existing antibiotics. As AMR continues to escalate, embracing the complexity inherent in nature may prove one of our most effective responses.

2. Materials and Methods

2.1. Data Sources and Literature Selection

This systematic review and meta-analysis were conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines (Page et al., 2021). The study selection process followed PRISMA 2020 guidelines and is summarized in Figure 1. A comprehensive systematic review and meta-analysis were conducted to evaluate the antimicrobial efficacy of plant-derived compounds and their combinatorial effects against clinically relevant pathogens. Peer-reviewed articles were retrieved from multiple electronic databases, including PubMed, Web of Science, Scopus, and Google Scholar, covering the period from 1999 to 2023. The search strategy combined Medical Subject Headings (MeSH) and free-text terms such as “plant-derived antimicrobials,” “phytochemicals,” “synergistic combinations,” “multidrug-resistant bacteria,” and “Cannabis-derived compounds.” Only studies reporting in vitro and in vivo antimicrobial activity with clearly defined minimum inhibitory concentration (MIC) values were included. Both single-compound and combination studies were considered, focusing on Gram-positive and Gram-negative bacterial pathogens, including the ESKAPE group. Exclusion criteria encompassed studies lacking quantitative antimicrobial data, non-English publications, abstracts without full text, and reports without standardized experimental protocols. Two independent reviewers conducting the initial screening and full-text assessment. Discrepancies were resolved through consensus or consultation with a third reviewer. Data extraction templates were developed in Microsoft Excel to collect study characteristics, compound types, microbial strains, MIC values, exposure times, combination ratios, and observed interaction types (synergistic, additive, or antagonistic).

2.2. Plant-Derived Compounds and Combinatorial Strategies

Selected studies evaluated various plant-derived secondary metabolites, including flavonoids, polyphenols, terpenes, cannabinoids, bitter acids, and essential oils. Specific compounds frequently reported included xanthohumol, lupulone, CBD, 7-hydroxyflavone, and CO2 extracts, as well as crude plant extracts rich in bioactive molecules. Combinations assessed in the included studies consisted of: (i) plant–plant interactions, (ii) plant–antibiotic interactions, and (iii) multi-component mixtures within a single plant extract. Interaction types were classified based on Fractional Inhibitory Concentration (FIC) index calculations or isobologram analyses, following established protocols. Synergy was defined as FIC = 0.5, additive interactions ranged from 0.5 to 1.0, non-interactive effects ranged from 1.0 to 4.0, and antagonism was considered at FIC > 4.0. Extraction methods were documented, including solvent-based maceration, Soxhlet extraction, CO2 supercritical extraction, and ethanol or methanol fractionation. Quality control measures ensured reproducibility, including reporting of solvent purity, plant part used, drying conditions, and storage parameters. Where reported, chemical characterization of extracts or compounds was verified using high-performance liquid chromatography (HPLC), nuclear magnetic resonance (NMR), or mass spectrometry (MS), confirming the presence of target bioactive constituents.

2.3. Microbial Strains and Antimicrobial Assays

Microbial targets included both standard laboratory strains and clinically isolated multidrug-resistant bacteria. Gram-positive strains primarily included Staphylococcus aureus (methicillin-sensitive and methicillin-resistant strains), Enterococcus faecalis, and Clostridium difficile, while Gram-negative targets comprised Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Bacteroides fragilis. Antimicrobial activity was assessed using standard broth microdilution methods to determine MIC values, following Clinical and Laboratory Standards Institute (CLSI) guidelines. For combination studies, checkerboard assays were employed to evaluate interactions, with FIC indices calculated for each combination. Time-kill assays were performed in selected studies to confirm synergistic effects over 24- to 48-hour exposure periods. Standard controls included untreated bacterial cultures, vehicle controls, and antibiotics with known efficacy. Replicates ranged from three to six independent experiments per study to ensure statistical validity. Data from each experiment were extracted, including mean MIC values, 95% confidence intervals, and standard errors. Where multiple trials of the same compound against the same pathogen existed, meta-analytic techniques were applied to synthesize results and calculate pooled estimates. Funnel plots were constructed to assess publication bias, with effect sizes plotted against precision metrics such as standard error or sample size.

2.4. Statistical Analysis and Data Synthesis

All quantitative data were synthesized using meta-analytic approaches appropriate for continuous outcomes, specifically MIC values expressed in µg/mL. Mean MIC values, standard errors, and 95% confidence intervals were extracted or calculated for each study. Random-effects models were employed to account for inter-study variability due to differences in extraction methods, pathogen strains, and experimental conditions. Heterogeneity was assessed using the I² statistic, with values above 50% indicating moderate to high heterogeneity. Subgroup analyses were conducted to evaluate differences in efficacy between compound classes, pathogen groups, exposure times, and interaction types (synergistic, additive, antagonistic). Sensitivity analyses were performed by excluding outlier studies or studies with extreme MIC values to assess the robustness of pooled estimates. Statistical significance was determined at a p-value < 0.05. Graphical representations, including forest plots, funnel plots, and bubble plots, were generated using R software (version 4.3) with the ‘meta’ and ‘metafor’ packages. Descriptive statistics complemented meta-analytic results to provide a comprehensive overview of antimicrobial efficacy across compound classes. Additionally, mechanistic insights were integrated by summarizing reported modes of action, including membrane disruption, efflux pump inhibition, enzyme interference, and synergistic potentiation of conventional antibiotics. This integrative methodology allowed for a systematic evaluation of the efficacy and safety potential of plant-derived compounds and their combinations against multidrug-resistant pathogens.

3. Results

The systematic aggregation and meta-analysis of plant-derived antimicrobial compounds revealed significant variations in potency across different chemical classes, pathogen groups, and exposure durations. Table 1 summarizes the mean minimum inhibitory concentration (MIC) values for major compounds and extracts against selected pathogens. Among the evaluated compounds, lupulone exhibited the highest potency against Staphylococcus aureus, with a mean MIC of 0.76 µg/mL (95% CI: 0.38–1.15) over 24 hours (Table 1). In comparison, CO2-extracts and xanthohumol demonstrated moderate activity, with mean MICs of 6.32 µg/mL (95% CI: 2.21–10.43) and 6.53 µg/mL (95% CI: 3.05–10.01), respectively. Flavonoid mixtures showed relatively lower activity, with a mean MIC of 22.64 µg/mL (95% CI: 12.42–32.85) against Gram-positive pathogens, highlighting that complex mixtures may require higher concentrations to achieve inhibitory effects. These differences reflect inherent chemical diversity and the variable capacity of individual compounds to permeate bacterial cell walls or interfere with critical cellular pathways.Figures, which graphically represent the grouped compound data, confirm the observed trends in Tables, illustrating that lupulone maintains consistent efficacy across replicates, whereas flavonoids exhibit broader variance, likely due to the heterogeneity of their constituent molecules. The relatively narrow confidence intervals for lupulone indicate reproducibility and robustness of effect, whereas wider intervals for flavonoids suggest possible variability in extract composition or differences in assay conditions. Notably, the 48-hour data for bitter acids against Clostridioides difficile revealed a substantially higher MIC of 388.0 µg/mL (95% CI: 274.54–501.46), emphasizing that some compounds, particularly those targeting anaerobic pathogens, may require extended exposure or higher doses to achieve bacteriostatic or bactericidal effects. Similarly, xanthohumol against Bacteroides fragilis at 48 hours exhibited a mean MIC of 39.43 µg/mL (95% CI: 27.80–51.06), suggesting time-dependent enhancement of antimicrobial activity.

Analysis of individual studies in Table 2 provides further insights into compound-specific performance against particular bacterial strains. For instance, Martinenghi et al. reported a mean MIC of 1.0 µg/mL for cannabidiol (CBD) against methicillin-resistant Staphylococcus aureus (MRSA), with a standard error of 0.25 µg/mL, indicating strong antimicrobial potential at low concentrations. Similarly, Math et al. reported a mean MIC of 1.6 µg/mL for 7-hydroxyflavone against S. aureus, further confirming the potency of flavonoid derivatives when assessed in isolation. In contrast, observed extremely high MIC values (2520 µg/mL) for E. coli treated with Cinnamomum zeylanicum essential oil, reflecting intrinsic resistance of Gram-negative pathogens to hydrophobic plant compounds, possibly due to the protective outer membrane limiting compound uptake.

The distribution of effect sizes plotted highlights notable variability across individual studies. While compounds like CBD, lupulone, and CO2 extracts cluster around low MIC values, essential oils and dimeric naphthopyranones exhibit dispersed effect sizes, reflecting both pathogen-specific susceptibility and methodological differences in extraction or assay protocols. Data visualizes individual study MIC values against standard error, suggests minimal publication bias, as the scatter appears relatively symmetrical around the weighted mean, though some studies with extreme MICs, represent potential outliers. These outliers, however, are biologically informative, demonstrating that Gram-negative pathogens generally require higher doses or combinatorial approaches to overcome innate resistance mechanisms.

A comparative assessment of compound classes reveals that prenylated chalcones (xanthohumol) and bitter acids demonstrate moderate to high activity against Gram-positive bacteria, particularly S. aureus, whereas flavonoid-rich extracts exhibit broader-spectrum but less potent effects. The superior efficacy of lupulone is likely attributable to its lipophilic nature, facilitating integration into bacterial membranes and disruption of membrane integrity. CBD and other Cannabis-derived compounds act synergistically with conventional antibiotics, as indicated by low MIC values in combination studies, suggesting an adjuvant effect that enhances antimicrobial potency while reducing the required concentration of each agent.

Temporal analysis, using 24- and 48-hour exposure data, indicates that most compounds display either time-dependent or concentration-dependent inhibition patterns. For example, xanthohumol’s MIC against B. fragilis at 48 hours (39.43 µg/mL) is substantially higher than against S. aureus at 24 hours (6.53 µg/mL), highlighting pathogen-specific kinetics. Bitter acids, in particular, demonstrate markedly high MICs against C. difficile at 48 hours, indicating that anaerobic pathogens may require extended exposure for effective inhibition. The FIC-index-derived classification of interactions, though not explicitly shown in the figures, aligns with these observations, as low MIC values in combinatorial studies correspond to synergistic interactions that potentiate antimicrobial activity.

Table 1. Grouped MIC Values by Compound/Extract. This table presents the mean minimum inhibitory concentration (MIC) values of high-potency compounds and extracts against specific bacterial groups. 95% confidence intervals (CI) are included to indicate variability across trials, enabling comparison of relative antimicrobial potency.

Table 2. Individual Study MIC Values. This table reports MICs for individual studies, including standard errors (SE), which can be used for funnel plots or meta-analytic forest plots to assess study precision and variability.

The meta-analytic synthesis supports the hypothesis that plant-derived compounds, either as single entities or in synergistic combinations, can substantially reduce bacterial growth, especially among multidrug-resistant Gram-positive strains. Grouped data illustrate that both extraction method and compound class significantly influence antimicrobial potency, with CO2 extracts and purified compounds generally outperforming crude mixtures. Individual study analyses highlight the importance of chemical standardization, as variability in MIC values among flavonoid and essential oil studies underscores the necessity of rigorous characterization and quality control for reproducible therapeutic effects.

Graphical analyses also provide evidence for selective targeting of pathogens. While Gram-positive bacteria consistently show low MICs across multiple compounds, Gram-negative bacteria like E. coli exhibit marked resistance, suggesting that combinatorial approaches, including adjuvant-mediated efflux pump inhibition, may be required to restore susceptibility. The distribution of standard errors in further emphasizes experimental robustness; low SE values for CBD and 7-hydroxyflavone indicate reliable assay conditions, whereas high SE values for essential oils reflect methodological heterogeneity.

Overall, the results underscore three primary trends: (i) prenylated chalcones and bitter acids are highly effective against Gram-positive pathogens, (ii) Cannabis-derived compounds and other plant secondary metabolites act synergistically with antibiotics to enhance efficacy, and (iii) Gram-negative bacteria require higher concentrations or combinatorial strategies for meaningful inhibition. Figures 1–4 collectively provide a visual summary of these findings, highlighting both pooled trends and individual study variation. The integration of grouped and individual study data ensures a comprehensive interpretation, demonstrating that plant-derived compounds hold considerable promise as standalone or adjunctive antimicrobial therapies, particularly in the context of multidrug-resistant pathogens.

3.1 Interpretation and discussion of the funnel and forest plots

The forest and funnel plots provide critical insights into the distribution of effect sizes, heterogeneity, and potential publication bias in the meta-analysis of plant-derived antimicrobial compounds. The forest plot (Figure 2) illustrates the individual study effect sizes alongside the pooled estimates for minimum inhibitory concentration (MIC) values across diverse pathogens. Each horizontal line represents the 95% confidence interval (CI) of a study, with the square denoting the point estimate of MIC, and the diamond representing the overall pooled effect.

Table 3. Antimicrobial Potency of Selected Extracts. This table summarizes the mean MIC (mg/mL) of various natural extracts against bacterial pathogens. 95% confidence intervals (CI) are included to indicate variability in reported activity, allowing comparison across compounds for forest plot representation.

Grouped MIC values and confidence intervals used for forest plot visualization are summarized in Table 3. A careful examination reveals that the majority of studies cluster around the pooled mean, indicating general agreement in the antimicrobial potency of certain compounds, particularly prenylated chalcones such as xanthohumol and bitter acids against Gram-positive bacteria like Staphylococcus aureus. Grouped compound-level trends in antimicrobial potency are visually summarized in Figure 2. The forest plot also highlights variability among specific studies, particularly those evaluating flavonoid mixtures and essential oils, where wider confidence intervals suggest differences in compound composition, extraction methods, or experimental conditions. This heterogeneity, reflected in both the width of the CIs and the degree of overlap, underscores the challenges in standardizing plant-based extracts for reproducible antimicrobial activity.

Importantly, the forest plot reveals that some studies report effect sizes that deviate significantly from the pooled mean, indicating potential outliers or highly specific pathogen–compound interactions. For example, studies examining E. coli or Clostridioides difficile show markedly higher MIC values compared to other Gram-positive strains, emphasizing inherent differences in bacterial susceptibility. These findings align with known biological mechanisms, as Gram-negative bacteria possess an outer membrane that restricts the entry of hydrophobic phytochemicals, whereas Gram-positive bacteria are more readily permeabilized. The forest plot’s overall pooled estimate, however, provides a reliable summary, demonstrating that, despite inter-study variability, plant-derived compounds exhibit measurable antimicrobial effects that are both statistically significant and biologically relevant.

The funnel plot (Figure 3) serves as a complementary tool to assess potential publication bias. In an unbiased dataset, studies with smaller sample sizes should scatter widely at the bottom of the funnel, while larger studies converge near the top, producing a roughly symmetrical, inverted funnel shape.

Table 4. MIC Values of Individual Compounds Against Specific Pathogens. This table provides mean MIC values with standard errors for individual bioactive compounds against specific bacterial targets. The data allow calculation of weighted effect sizes for meta-analysis or forest plot visualization.

The funnel plot in this analysis appears largely symmetrical, suggesting minimal publication bias, with a balanced distribution of small- and large-sample studies on either side of the pooled effect size. Some asymmetry is noted for studies assessing essential oils and complex flavonoid extracts, which may reflect selective reporting or variability in methodological rigor, rather than systematic bias. Notably, the symmetry observed for key compounds like lupulone and xanthohumol indicates that their reported MIC values are robust and reproducible across multiple independent investigations.

The combination of forest and funnel plot analyses also highlights the reliability of the pooled estimates derived from the meta-analysis. The forest plot confirms that the majority of effect sizes fall within narrow confidence intervals, particularly for single-compound studies, suggesting high precision in these measurements. Meanwhile, the funnel plot reinforces that this precision is not an artifact of selective publication, strengthening confidence in the validity of the overall findings. This dual assessment is particularly important in phytochemical research, where variability in extraction, concentration, and assay protocols can otherwise confound interpretation.

Furthermore, the plots allow for identification of trends and potential gaps in the literature. The forest plot indicates that Gram-positive pathogens consistently exhibit lower MIC values across multiple compounds, whereas Gram-negative bacteria show both higher MIC values and wider variability. This pattern suggests that future research may need to focus on combinatorial or adjuvant strategies to enhance the efficacy of plant-derived compounds against more resistant Gram-negative strains. Similarly, the funnel plot indicates a relative paucity of large-scale studies for certain compound–pathogen combinations, particularly flavonoid-rich extracts, highlighting areas where additional rigorous experimentation is warranted to confirm preliminary observations.

Finally, integrating insights from both plots provides a nuanced understanding of the overall antimicrobial landscape. The forest plot quantifies the effect and its heterogeneity, identifying both effective compounds and experimental outliers. The funnel plot, by assessing symmetry and dispersion, evaluates the credibility and potential biases of the dataset. Together, they suggest that the observed antimicrobial activities are genuine and reproducible, rather than artifacts of selective reporting, while also guiding future research priorities, including standardization of extraction methods, targeted evaluation against resistant Gram-negative pathogens, and rigorous replication of studies with complex phytochemical mixtures.

In conclusion, the forest and funnel plots collectively confirm the robustness of the meta-analytic findings, illustrating consistent antimicrobial efficacy for several key compounds, highlighting sources of heterogeneity, and revealing minimal publication bias. These analyses underscore the potential of plant-derived compounds as effective antimicrobial agents, particularly against Gram-positive pathogens, while also identifying critical areas for methodological improvement and further research.

4. Discussion

The present meta-analysis highlights the considerable potential of plant-derived compounds and extracts as alternative antimicrobial agents, particularly in the context of rising multidrug-resistant pathogens. The integration of systematic review data with meta-analytic synthesis revealed both consistent antimicrobial efficacy and notable variability across studies, emphasizing the complexity inherent in phytochemical research. The observed MIC values from Tables 1 and 2, alongside forest and funnel plot analyses (Figures 3, and 5), underscore that specific compounds, such as lupulone, xanthohumol, and certain flavonoids, consistently exhibit strong activity against Gram-positive bacteria like Staphylococcus aureus, whereas Gram-negative strains such as E. coli and Bacteroides fragilis display higher MICs and greater variability. These trends align with established biological principles, given the structural differences in bacterial cell envelopes that influence compound permeability and susceptibility (Vaou et al., 2022; Mickymaray, 2019).

The forest plots (Figures 2 and 4) indicate that most individual study effect sizes cluster near the pooled mean, particularly for high-potency agents such as lupulone and xanthohumol, suggesting reproducibility and robustness in their antibacterial activity. This convergence contrasts with the broader confidence intervals observed in flavonoid-rich extracts and essential oils, highlighting the inherent variability of complex plant mixtures (van Vuuren & Viljoen, 2011; Cowan, 1999). Such variability likely arises from differences in phytochemical composition due to extraction methods, plant source, or environmental factors, illustrating the importance of standardized protocols for both preclinical and clinical evaluation (Wagner & Ulrich-Merzenich, 2009; Spelman et al., 2006).

Synergistic interactions, as discussed in the Introduction, appear to play a pivotal role in enhancing antimicrobial efficacy. The fractional inhibitory concentration (FIC) indices and literature evidence indicate that combinations of cannabinoids, polyphenols, and bitter acids can potentiate the activity of conventional antibiotics (Karas et al., 2020; Hemaiswarya et al., 2008; Wassmann et al., 2020). For example, cannabidiol (CBD) demonstrated robust synergy with bacitracin and polymyxin B against resistant Gram-positive strains, effectively lowering the required MIC and reducing potential toxicity (Farha et al., 2020; Blaskovich et al., 2021). Similarly, flavonoid-rich extracts from Mangifera indica and Eryngium campestre not only exert direct antimicrobial effects but also enhance the activity of standard antibiotics by disrupting bacterial defense mechanisms such as efflux pumps (Yehia & Altwaim, 2023; Al-Askar et al., 2023). These findings reinforce the significance of multi-component approaches, supporting the notion that the collective effect of plant-derived compounds often exceeds the sum of individual activities (Wojtyczka et al., 2013; Hemaiswarya et al., 2008; Vaou et al., 2022).

The analysis of funnel plots (Figure 3) provided additional confidence in the validity of these findings, revealing minimal asymmetry for key compounds such as lupulone and xanthohumol, suggesting that reported MICs are unlikely to be biased by selective publication. Some asymmetry was noted for essential oils and flavonoid combinations, which may reflect methodological heterogeneity rather than systematic bias (Langeveld et al., 2014; Bassolé & Lamien-Meda, 2010). The overall symmetry reinforces the reliability of the meta-analytic estimates, indicating that high-potency compounds consistently demonstrate antimicrobial activity across independent studies.

Several studies included in this analysis also underscore the importance of integrating computational and in vivo approaches with traditional antimicrobial assays. For instance, virtual screening and pharmacophore modeling have been used to identify plant-derived inhibitors targeting essential bacterial proteins, such as Hsp90, demonstrating the potential of bioinformatics-guided discovery in natural product research (Abbasi et al., 2021; Ouassaf et al., 2023). Moreover, in vivo validation using Caenorhabditis elegans models allows for the assessment of compound safety, efficacy, and pharmacodynamics in a whole-organism context, bridging the gap between in vitro assays and higher vertebrate studies (Moy et al., 2006; Zarroug et al., 2023). Such integrative approaches are critical for translating phytochemical discoveries into clinically relevant therapies.

Despite the promising results, this meta-analysis revealed substantial heterogeneity among certain compound classes, particularly complex flavonoid and essential oil extracts. The wide confidence intervals and variability in MICs suggest that minor differences in preparation, concentration, or assay conditions can profoundly influence antimicrobial outcomes (van Vuuren & Viljoen, 2011; Eloff, 2004). These findings highlight the necessity of rigorous standardization in future studies, including consistent plant sourcing, extraction protocols, and validated microbial assays, to ensure reproducibility and accurate assessment of pharmacological potency. Furthermore, understanding the mechanistic basis of synergistic interactions remains a crucial research priority, as elucidating the molecular targets and pathways modulated by compound combinations can guide rational formulation of multi-component therapeutics (Stermitz et al., 2000; Junio et al., 2011; Farha et al., 2020).

The results also emphasize pathogen-specific considerations. While Gram-positive bacteria consistently exhibited low MICs for key phytochemicals, Gram-negative pathogens displayed markedly higher MICs, reflecting intrinsic resistance mechanisms such as the outer membrane barrier and efflux pumps (Cermak et al., 2023; Durães et al., 2021). These findings suggest that plant-derived combinations may be particularly useful as adjuvants for Gram-negative infections, enhancing the potency of existing antibiotics while circumventing conventional resistance pathways. In this context, the role of synergistic plant compounds as resistance-modifying agents becomes highly relevant (Rather et al., 2013; Hemaiswarya et al., 2008).

Finally, the forest and funnel plot analyses collectively highlight both the reproducibility of high-potency compounds and the need for further studies on complex extracts and Gram-negative pathogens. The consistency of effect sizes for lupulone, xanthohumol, and CBD demonstrates their translational potential as adjunctive therapies, while the variability observed in flavonoid-rich and essential oil preparations identifies targets for future investigation. Integrating systematic review data with meta-analytic synthesis thus provides a robust framework for prioritizing compounds with the highest clinical promise while informing experimental design and mechanistic studies in natural product research. Study-level MIC values used for weighted meta-analytic calculations are detailed in Table 4. In conclusion, this meta-analysis supports the substantial antimicrobial potential of plant-derived compounds, particularly when employed in synergistic combinations. High-potency agents such as lupulone, xanthohumol, and cannabinoids demonstrate reproducible activity against Gram-positive pathogens, while flavonoid-rich and essential oil extracts offer broader, albeit more variable, antimicrobial effects. The combination of forest and funnel plot analyses confirms the reliability of these findings, minimizes concerns regarding publication bias, and provides a foundation for the rational development of multi-component phytotherapeutics. Future research should focus on standardization of extraction methods, mechanistic elucidation of synergy, and exploration of combinatorial strategies to overcome Gram-negative resistance, thereby contributing to the development of effective, natural alternatives in the global fight against antimicrobial resistance.

5. Limitations

Despite the comprehensive synthesis of plant-derived antimicrobial combinations in this study, several limitations must be acknowledged. First, the reliance on published data may introduce publication bias, as studies reporting significant synergistic effects are more likely to be published than those reporting null or antagonistic outcomes. Second, the heterogeneity of experimental methodologies across studies, including variations in extraction methods, compound concentrations, assay types, and incubation periods, may influence the reported MIC values and interaction classifications, potentially limiting direct comparability. Third, many studies utilized in vitro models, which, while informative, may not fully capture the complex pharmacokinetic and pharmacodynamic interactions occurring in vivo, thereby affecting the translational relevance of the findings. Fourth, while statistical analyses such as pooled MIC values, forest plots, and funnel plots were employed, the relatively small number of studies for some compounds may reduce the robustness of meta-analytic conclusions and increase susceptibility to random errors. Additionally, interactions between compounds are highly context-dependent, and environmental factors, microbial strain variability, and bioavailability in real-world applications were not fully accounted for. Finally, the focus on high-potency compounds may overlook the potential synergistic contributions of minor constituents, which are integral to the holistic effects of medicinal plant extracts. These limitations suggest cautious interpretation and the need for standardized experimental protocols and in vivo validation.

6. Conclusion

This study highlights the potent antimicrobial potential of plant-derived combinations, emphasizing synergistic interactions that enhance efficacy while reducing resistance and toxicity. Statistical analyses support the superior activity of complex mixtures over single compounds against pathogenic bacteria, including resistant strains. These findings underscore the value of integrating phytochemicals into antimicrobial strategies. Future research should prioritize standardized methodologies, in vivo validation, and exploration of minor constituents to fully harness the therapeutic potential of medicinal plants in combating antimicrobial resistance.

Author Contributions

S.F.J. conceptualized the study, designed the review framework, and drafted the original manuscript. S.G. conducted literature analysis, contributed to data synthesis, and assisted in manuscript preparation. M.N. contributed to data interpretation, methodological support, and critical analysis. F.A. contributed to manuscript revision, intellectual input, and final editing.  All authors read and approved the final version of the manuscript.

References