Antibacterial activity on multidrug resistance bacteria stenotrophomonas maltophilia (strain k279a) and its correlation to secondary metabolic profile of heavy metal hyperaccumulator pteris vittata l

UNIVERSITY OF SCIENCE AND TECHNOLOGY OF HANOI UNIVERSITÉ DES SCIENCES ET DES TECHNOLOGIES DE HANOI UNDERGRADUATE SCHOOL FINAL REPORT BACHELOR GRADUATION By Tô Hiền Minh Project: Ant

UNIVERSITY OF SCIENCE AND TECHNOLOGY OF HANOI UNIVERSITÉ DES SCIENCES ET DES TECHNOLOGIES DE HANOI

UNDERGRADUATE SCHOOL

FINAL REPORT BACHELOR GRADUATION

By

Tô Hiền Minh

Project:

Antibacterial activity on multidrug resistance bacteria

Stenotrophomonas maltophilia (strain K279a) and its

correlation to secondary metabolic profile of heavy metal

hyperaccumulator Pteris vittata L

University of Science and Technology of Hanoi

Hanoi, November 2020

Index of Contents

Contents

Acknowledgements

List of Abbreviations

List of Figures and Tables

Abstract:

Tóm tắt:

I Introduction: 1

1.1 Global context: 1

1.2 Literary review: 4

1.3 Objective: 8

II Materials and Methods: 8

2.1 Chemicals 8

2.2 Instruments 9

2.3 Biological materials 9

2.4 Preparation of plant extracts 9

2.5 Bacterial and growth conditions 9

2.6 Antibacterial activity determination 10

2.7 Metabolite profiling of the methanolic extracts of P vittata 10

2.8 Data analysis 10

III Results: 12

3.1 Antibacterial effect of Pteris Vittata roots and leave extracts on resistance strain Stenotrophomonas maltophilia K279a: 12

3.2 Correlation between antibacterial activity of 70 Pteris Vittata extracts and its metabolic profile: 17

IV Discussion: 21

4.1 Discrimination on the activity of 70 Pteris vittata roots and leave extracts on S.malophilia strain K279a: 21

4.2 Correlation analysis between metabolic profile and antibacterial activity of Pteris Vittata 23 V Conclusion: 25

VI Reference: 26

VII Appendices

Acknowledgements

Foremost, I would like to express my deepest gratitude to my supervisors Dr Nguyễn Thị Kiều Oanh and PhD student Nguyễn Ngọc Liên from Life Sciences Department for their countless support, encouragement and guidance throughout my 3- month internship in the University of Science and Technology Hanoi I could not have broadened my research skills and biostatistics knowledge without the opportunity that they give me to join the project and fulfill my bachelor thesis

Besides, I want to thank all of the other professors and lecturers in the Life Science Department who always inspire us with their enthusiasm and passion for science - the best role models for us to become in the future Moreover, I’m sincerely proud to spend my most wonderful 3 years of education with all of my friends in the PMAB class of 2020 We have always tried hard and inspired each other to break our own limits Especially Phạm Anh Thư and Trần Phương Trang for always supporting and motivating me The final semester

I had an opportunity to do an exchange program in Italy thanks to our university with a wonderful international cooperation strategy

Most importantly, all of my thankfulness goes to my family and loved ones, they have nurtured and educated me to become who I am today, who is not afraid to learn and always grateful for what is coming in the future

List of Abbreviations

P vittata Pteris vittata L

S maltophilia Stenotrophomonas maltopilia

TMP-SMX trimethoprim- sulfamethoxazole

MRSA Methicillin-resistant Staphylococcus aureus

UPLC-QqQ-MS Ultra Performance Liquid Chromatography coupled triple

quadrupole mass spectrometry

PLS Partial least square regression

PLS- DA Partial least square regression- Discrimination analysis

List of Figures and Tables

Figures

Figure 1 Microdilution testing of P vittata root and leave extract on the growth curve of

S maltophilia strain K279a

Figure 2 The inhibition percentage of P vittata leave and roots extracts on the growth of

S maltophilia strain K279a

Figure 3 Correlation circles of PLS regression between metabolites content (M) and the

inhibition percentage of S maltophilia K279a growth by leave and root extracts (Y)

Tables

Table I Paired t-Test statistic result on the difference between leave and root extracts

activity

Table II Pearson correlation coefficient statistics measure the linear correlation between

metabolite intensity (X1) and antibacterial activity of leave or root extracts (X2)

Abstract:

Pteris vittata L is a heavy metal hyperaccumulator plant abundant in Thai Nguyen

mining ore of Vietnam, which can change its secondary metabolic profile for environmental adaptation, leading to the shift of its synergistic bacterial system Under the same heavy metal stress, multidrug resistance bacteria can express intrinsic efflux pumps for both metal

and antibiotics extrudation Investigation on P vittata bioactivity is a potential approach in

finding novel antimicrobial agents to cope with an emerging era of multidrug resistance

bacteria The resistant bacteria Stenotrophomonas maltophilia strain K279a is an

opportunistic nosocomial pathogen which can cause severe respiratory tract infection

Microdilution testing shows that P vittata roots and leave extracts have either antibacterial

activity or growth promoting activity on the strain In our study, the first time metabolic

profile of P vittata identified by a UPLC-QqQ-MS was statistically predicted for its

correlation with antibacterial activity Flavonoid compounds such as pelargonin, quercetin

were predicted to act against the S maltophilia and some alkaloids show its growth

stimulator role These results bring out new directions to develop antibacterial lead

compounds for treating clinical infection of Stenotrophomonas maltophilia K279a as well

as to study its new resistance tendency

Keywords: Pteris vittata, Stenotrophomonas maltophilia, antibacterial activity, bacterial

growth promoting, secondary metabolites, PLS regression model

Tóm tắt:

Pteris vittata L là một thực vật hấp thụ kim loại nặng xuất hiện rất nhiều tại vùng

mỏ ở Thái Nguyên, Việt Nam Nó có thể thay đổi các chất chuyển hóa thứ cấp của mình để thích nghi với môi trường, dẫn đến sự thay đổi của hệ vi khuẩn sống cộng sinh với nó Trong cùng một môi trường kim loại nặng, vi khuẩn đa kháng thuốc hình thành các bơm tống có chức năng đào thải cả kim loại nặng và thuốc kháng sinh Việc nghiên cứu hoạt

tính sinh học của P vittata là một cách tiếp cận tiềm năng để tìm ra các chất kháng khuẩn

và đối phó với thời đại vi khuẩn đa kháng thuốc đang nổi lên hiện nay Vi khuẩn đa kháng

thuốc Stenotrophomonas maltophilia chủng K279a là một mầm bệnh cơ hội trong bệnh

viện có thể gây nhiễm trùng đường hô hấp nặng Thử nghiệm pha loãng đa nồng độ cho

thấy dịch chiết từ rễ và lá của P vittata có hoạt tính kháng khuẩn hoặc hoạt tính thúc đẩy

tăng trưởng trên chủng Trong nghiên cứu của chúng tôi, lần đầu tiên bộ chất chuyển hóa

của P vittata phân lập bởi UPLC-QqQ-MS được dự đoán thống kê về mối tương quan của

nó với hoạt tính kháng khuẩn Các hợp chất flavonoid như pelargonidin, quercetin được dự

đoán là có tác dụng chống lại S maltophilia và một số alkaloid cho thấy vai trò kích thích

tăng trưởng trên chủng Những kết quả này đã mở ra hướng mới trong việc phát triển các

hợp chất kháng khuẩn để điều trị nhiễm trùng lâm sàng bởi S maltophilia K279a cũng như

nghiên cứu xu hướng kháng thuốc mới của chủng vi khuẩn này

Từ khóa: Pteris vittata, Stenotrophomonas maltophilia, hoạt tính kháng khuẩn, hoạt tính

tăng trưởng, thành phần chất chuyển hóa, hồi quy bình phương tối thiểu

1

1.1 Global context:

Antibiotic resistance situation worldwide

Antibiotic resistance is becoming an urgent public health concern worldwide Human opportunistic pathogens cause millions of people with related diseases and thousands of deaths each year While becoming a burden for drug investigating and post-surgery infections prevention, the resistance strains show no sign of weakening (CDC, 2020) Reason for antibiotic resistance is not only the rapid evolving of bacteria themselves but also from anthropogenic activity: over-prescribing of antibiotics, excessive antibiotics used in agriculture and poor hygiene leading to nosocomial infections in hospital environments Others worth mentioning are the lack of rapid laboratory tests and global record of bacteria resistance situations

In general, the most common nosocomial resistant bacteria strains are Pseudomonas aeruginosa, Stenotrophomonas maltophilia, Klebsiella pneumoniae, Staphylococcus aureus, Streptococcus pneumonia and Mycobacterium tuberculosis Report in 2018 shows approximately 640.000 cases of multidrug-resistant tuberculosis disease While penicillin

as antibiotic against gram-positive bacteria experienced a trend of weakening in the ability

to treat pneumonia, S maltophilia associated with respiratory tract infections also counters

against its commonly used inhibitor trimethoprim- sulfamethoxazole (TMP-SMX)(“WHO

| High Levels of Antibiotic Resistance Found Worldwide, New Data Shows” n.d.) In a worse fact, some strains become multi-resistance to any available clinical antimicrobial

agents such as Methicillin-resistant Staphylococcus aureus(MRSA) which is notorious for

hospital-acquired infections and resistant to methicillin, aminoglycosides, macrolides, tetracycline, vancomycin along with disinfectants Similarly, the occurrence of pan-

resistant bacteria such as Pseudomonas aeruginosa, Acinetobacter baumannii, S

maltophilia which conquers all of the antibiotics in standard panels via the mechanism of

multidrug efflux pumps (Nikaido 2009)

2

Recently, Covid-19 situation has also aggravated antibiotic resistance Statistics show that 50% of Covid-19 mortalities suffered from secondary infections, caused by multidrug resistance microorganism (bacterial or fungal) or coinfections Considering that these patients have already had underlined bacterial or fungal risk diseases (e.g: corticoid therapy, chronic respiratory disease, immune inflammatory response) and responded to a much higher rate of antibiotic usages when they are hospitalized (Rossato, Negrão, and Simionatto 2020)

Plant as a novel source for producing antibiotics

The bacteria possess resistance to multiple drug thanks to either one of these two mechanisms: (i) bacteria accumulate multiple genes, each gene encodes for resistance to a single drug within a single cell, occur on resistance plasmids, (ii) bacteria increase expression of genes which code for multidrug efflux pumps extruding a wide range of drugs (Nikaido 2009) From the first day, antibiotics were identified by Alexander Flemming as

an accident leading to Penicillin invention Accordingly, the common sources of antibiotics come from secondary metabolites of fungi or soil bacteria produced by these industrial methods: fermentation (e.g: beta lactam, penicillins, cephalosporins); semi-synthetic; synthetic pathway In this scenario, the main source of antibiotics - soil actinomycetes has been overmined and the investigation based on this source is said to be useless due to the high rediscovery rate of known compounds (Lewis 2017) Rising as a novel candidate to conquer MDR bacteria, plant resources have long been used as herbal medicine to treat a vast array of inflammatory and bacterial infections It’s reported through archeology that in the primitive era, humans benefited partially from plant and metal to treat microbial infections: Neanderthals living 60,000 years ago in present-day Iraq used plants such as hollyhock to treat infections Since plants can produce a variety of primary and secondary metabolites, these compounds are not only active to defend insects, microorganisms, herbivores and plant pathogens but also against human pathogens The major groups are: phenolics, polyphenols, terpenoids, essential oils, lectins and polypeptides, alkaloids or polyacetylenes They can act solely or in combination with different compounds by a

3

variety of mechanisms: protein inactivation, adhesins and enzymes, block cell-to-cell signaling, quench production of virulence factors such as exotoxin, disrupt or inhibit formations of biofilms, shielding of pathogens during an infection (Cowan 1999) Despite such potentials, antimicrobial drugs from plants only consist of 3% drug approved by FDA Thus, the current investigation on antimicrobial potential of plants is merely the tip of the iceberg

Pteris vittata L antibacterial potential

Pteris vittata L., also known as Chinese Brake Fern, is well studied for its metal

hyperaccumulator function found in anthropogenic polluted ore mining sites in Thai Nguyen province of Vietnam A number of previous publications focus on its phytoremediation role, especially for the individual absorption of Pb, Zn, As and their combination For Arsenic, it can absorb up to 23,000 micro gram arsenic g-1 in its frond or aerial parts, 4-8 folds more efficiently than soil extraction method (Alkorta et al 2004) Otherwise, due to the phytotoxic correlation of excessive heavy metal and production of ,

P vittata is proved to have antioxidant activity by scavenging toxic free radicals and ROS’s

enzymatic or non enzymatic defense mechanism However, there is a limit of papers

investigating the antibacterial effect of P vittata or considering its roots and bacteria

synergism affected by the secondary metabolites While considering secondary metabolites

as strong antioxidants to effectively protect against oxidative stress, we should also put eyes

on the antibacterial potential of the diverse metabolic profile of P vittata

4

1.2 Literary review:

Antibiotic resistance in Vietnam:

Vietnam is one of the developing countries facing a severe stress of multidrug resistance due to the large population, high burden of infectious disease and the loose end

of prescription drugs Vietnam is ironically placed number one in penicillin and erythromycin resistance, at 71.4% and 92.1% resistance respectively, and also at 14th for

MDR-Tuberculosis in the world Hospital infection is also a great burden since the

expenditure for human and equipment to conquer is surmountable It’s recorded in 2001 that the rate of nosocomial infection is 6.8% and half of which causes pneumonia(Alp et al 2004) Accordingly, there are 3 most critical resistance strains involving in high rate

mortality nosocomial pneumonia patients: Pseudomonas aeruginosa, Acinetobacter spp.,

S maltophilia, all of which are multidrug resistant to beta-lactams, aminoglycosides and

fluoroquinolones (Ministry of Health, 2013)

Stenotrophomonas maltophilia is an aerobic nonfermentative Gram negative rod

bacillus commonly found in respiratory tract infections The bacteria are ubiquitous in aqueous habitats and found in both clinical and natural settings of plants, animals and water sources For years, it has emerged as an opportunistic nosocomial pathogen associated with high rate morbidity and mortality in immunocompromised patients (Brooke 2012) The

main clinical manifestation of S maltophilia infections can be listed as bacteremia,

endocarditis, urinary tract infections or pneumonia in cystic fibrosis patients The multidrug

resistance in S maltophilia emerged intrinsically due to the strain’s characteristic of low

membrane permeability and mechanism to produce multidrug resistance efflux pumps, lactamases, antibiotic-modifying enzymes and quinolone, aminoglycosides resistant genes

β-(Sánchez 2015) Consequently, S maltophilia shows a broad array resistance to its

commonly-used antibiotics: trimethoprim-sulfamethoxazole (TMP-SMX), β-lactam antibiotics, macrolides, cephalosporins, fluoroquinolones, aminoglycosides, chloramphenicol, tetracyclines, and polymyxins (Brooke, 2012) Worse still, it can acquire resistance genes horizontally through plasmids, transposons, integrons or sharing

5

exogenous genes within their rhizosphere bacterial community to advance its

pan-resistance ability Our focus strain of S maltophilia - K279a overexpresses nine pan-

resistance-nodulation-division (RND) type efflux pumps via the specific operon smeDEF, which causes hyper-resistance to fluoroquinolones, chloramphenicol and tetracycline (Crossman

et al 2008) In Vietnam, high prevalence of MDR S maltophilia was found in the food and

drink market (edible and conservative ice), hospital Intensive Care Units infections and rhizosphere soil environment (Dat et al 2017; Denet et al 2018) Resistance rate is very high at over 90% resistance to beta-lactams group (penicillins, cephalosporins, carbapenems); up to 100% resistance to aminoglycosides; 40% to fluoroquinolones, 30%

to tetracycline, 75% chloramphenicol (Son 2019) These carriages of multidrug resistance can be the reservoir for inter transferable between species, including the transfer of extended spectrum beta lactamase (ESBL), colistin-resistance gene and virulence gene

associated with other resistance bacteria The gradual evolving of MDR Stenotrophomonas maltophilia can cause severe problems to human health, challenges to investigate the new

treatment regime as well as huge economic burden to effectively conquer the side effects and invest in innovative pharmaceutical solutions

Pteris vittata L as a heavy metal hyperaccumulator

Due to exaggerating anthropogenic activities of mining, ores smelting, fuel production, fertilizer, pesticide, soil metal pollution is becoming a major environmental problem Heavy metal contamination of Ni, Cr, Pb, Cu, Se, Hg, Cd presents well above the permitted standards at hundreds of times in mining sites of Tuyen Quang, Thai Nguyen, Vietnam This leads to contamination in both soil and ground water for mining and agriculture cultivation, resulting in severe health problems Only 20% of the contaminated sites can be restored by traditional preliminary soil extraction technique, which is time and cost consuming (Anh and Kim 2011) Phytoremediation therapy is an aesthetic environmental-friendly, cost-effective strategy to extract, sequester and detoxify heavy

metal by plant intrinsic mechanism Pteris vittata L is proved to be a strong Arsenic

hyperaccumulator, accumulate up to 23.000 microgram arsenic g-1 in its fronds, has a

6

remarkable transportation of As from roots to shoots and accumulating up to 95% As in its

above ground tissue (Alkorta et al 2004) It can also sequester other heavy metal along

with As eg: K, P, Fe, Mn, Zn, Cu, Cd and produce non enzymatic (glutathione, acid-soluble thiol) as well as enzymatic antioxidant (superoxide dismutase, catalase, ascorbate peroxidase) to scavenge ROS produced by excessive heavy metals (Cao, Ma, and Tu 2004)

Pteris vittata’s secondary metabolites and soil microbial community correlation in

antibacterial activity:

In the heavy metal polluted area, soil bacterial communities have an ability to increase, emerge and spread their multi-drug resistance mechanism via the efflux pump mechanism which they express not only to pump out heavy metal but also antibiotics as organic compounds (Martinez et al 2009)

On one hand, S maltophilia strain K279a reveals strong capacity for heavy metal

resistance via their DNA mobile region (pili/fimbrial genes) and efflux protein, especially for arsenic, mercury and copper resistance (Crossman et al 2008) In 2018, Yunfu Gu

determined S maltophilia rhizospheres from P vittata arsenite accumulated ground with a

function as arsenic transporter and gene transfer, proving their strong cooperation in heavy

polluted conditions (Gu et al., 2018) On the other hand, P vittata has their own adaptation

mechanism to grow normally in metal stress and establish a strong synergism with

rhizosphere bacterial communities In particular, P vittata can change its own metabolic

profile, especially secondary metabolites to alter the soil bacterial growth as well as the MDR pump that they express The same project framework has identified a significant increase of chlorogenic acid derivatives and A-type procyanidin in contaminated derived plant roots, both have antibacterial activity (Pham et al 2017) While chlorogenic acid bind and exhaust bacterial outer membrane (Lou et al 2011), A-type proanthocyanidins have anti-adhesion activity to gram negative bacteria such as Pseudomonas aeruginosa (Ulrey

et al 2014) or uropathogenic E coli, prevent it from adhering to uroepithelial cells and

maintain urinary tract health (Howell et al 2005) Also, (Singh et al 2008) worked on antimicrobial activity against pathogenic gastrointestinal microflora, in which micro-

phyto activity of P vittata is shown by its antiproliferative action on human MCF-7 breast

cancer cells (Kaur et al 2014)

All together, we can hypothesize that P vittata, via their secondary metabolites, can give an influence on the acquisition/expression of MDR bacteria, especially with S maltophilia (strain K279a)

Metabolic profiling and correlation analysis by metabolomics-guided isolation

Previous stages in the same project framework use the technique UPLC-QqQ-MS (Ultra performance liquid chromatography-coupled triple quadrupole mass spectrometry)

to analyze the primary and secondary metabolites of P vittata This technique analysed 70

samples (including leave and roots of 35 plant individuals) through ultra high pressure liquid chromatography column, then mass-analyses them by a 4 quadrupole mass analyzers

in tandem with the mass spectrometry method, resulting in 359 metabolites identified The

correlation of this metabolic profile with antibacterial activity from P vittata microdilution

data was analyzed by PLS multivariate analysis Partial Least Square (PLS) is a dimensionality reduction supervised method developed from the Principal Component Regression method (PCR) with a similar function to reduce a large data set into a few principle components for multivariate correlation analysis This model works by projecting the measured chemical and spectroscopic variables on matrices X and Y to generate a linear regression for predicting their critical relationship In our PLS analysis, the relationship of extracts antibacterial activity and 359 metabolites is examined graphically (grouped by multiple components) The coordinates are calculated by the correlation of each variable (metabolites or percentage of bacterial inhibition) with the components Our variables are presented as the endpoint of the vectors and Y point is the antibacterial activity inhibition

8

percentage The product of the vector (angle of endpoint in our illustration) indicates the relationship between them (González et al 2012) PLS is more advanced than PCA as a robust and easy-visualizing method with more significant dimensions Even though it still has some limitations such as slower speed and requirement on minimum response effects (Hanrahan, Udeh, and Patil 2005)

1.3 Objectives:

Literary review shows a limited number of papers considering the antimicrobial activities

of P vittata in general and which secondary metabolites/bioactive compounds act against

these MDR bacteria

Therefore, this research will be focused on (i) Evaluate the antibacterial activity of

methanol extract of leave and roots of P vittata L against MDR S maltophilia (strain K279a) (ii) Correlation between metabolic profile and bacterial inhibition percentage and (iii) Propose potential active compound(s) against MDR S maltophilia K279a to be the

lead candidate(s) for drug development

II Materials and Methods:

2.1 Chemicals

For determining antibacterial activity, Soyabean Casein Digest Medium (Tryptone soya broth – TSB) was purchased from HIMEDIA (India) Agar- Agar was purchased from Merck (Darmstadt, Germany) DMSO (Dimethyl sulfoxide) reagent was purchased from Cica (Kanto, Japan) Chloramphenicol was obtained from National Institute of Drug Quality Control (Hanoi, Vietnam) Methanol was obtained from Aladdin (Shanghai, China) Distilled water and ultrapure water were prepared using a Millipore Milli-Q purification system (Millipore GmbH, Schwalbach, Germany)

35 individuals of P vittata samples were collected at different places in Hich village,

Tan Long ward, Dong Hy district, Thai Nguyen province, Vietnam (N: 21o43’612” E: 105o51’380”) under dry weather conditions Plant samples were identified by Dr Nguyen The Cuong, Institute of Ecology and Biological Resources, Vietnam Academy of Science and Technology The voucher of specimens was deposited in the Department of Life Science, University of Science and Technology of Hanoi

2.4 Preparation of plant extracts

For sample preparation, collected plants were separated into two parts: roots and leave The root part was washed in water to remove all dust and soil, then rinsed in distilled water All parts were cut down into small pieces, drying in the oven at 55°C for 48h until the weight does not vary Sample was extracted with 80% methanol/water (10 g/ 100 ml) at room temperature in a sonic bath, followed by centrifugation Extraction process was repeated in triplicate The combined extracts of 70 samples (including 35 roots and 35 leave) were concentrated in vacuo to obtain crude extracts and stored at −20 °C until analysis

2.5 Bacterial and growth conditions

The tested S maltophilia K279a strain was obtained from clinical samples (Brooke,

2012) This strain has a resistance pattern to TMP-SXM, polymyxin and colistin therapy Bacterial strain was grown in Tryptone soya broth medium (HIMEDIA, India), and incubated at 37oC with shaking for 16-18 h in incubator The optical density value approximate 0.4 was measured by spectrophotometry (equivalent to 10^9 CFU/mL) in TSB These suspensions were then diluted in TSB in order to obtain stock solutions containing 2.106 CFU/mL

10

2.6 Antibacterial activity determination

The antibacterial activity of crude extracts from P vittata dried roots and leaves were

evaluated by gold standard method micro-dilution assay (Esposito et al 2020; Siwe Noundou et al 2015) with some modification in accordance with the tested bacterial strain:

S maltophilia K279a Stock solutions of all the extracts (20.48 mg/ml) were prepared in

dimethyl-sulfoxide 100% (DMSO) Twentyfold dilution was made with TSB to reach concentration 1.024 mg/ml in DMSO 5% 100μl of each solution was first introduced to a 96-well microtiter plate Equal volumes (100 μl) of bacterial suspension grown overnight yielding an approximate inoculum size of 106 colony forming units (CFU)/mL in TSB was added to the sample wells with each blank well of the equal volumes (100 μl) of sterile TSB The final concentrations of each extract in wells was 512 μg/ml with a final concentration of DMSO at 2.5% Antibiotic (gentamicin 32 μg/ml) was used as positive control and DMSO at 2.5% in TSB was used as negative controls of growth Bacterial growth was indicated by determining the OD of the solution in each well at the wavelength

of 600 nm Wells were done in triplicate and data were then averaged

2.7 Metabolite profiling of the methanolic extracts of P vittata

In our previous study, UPLC-QqQ-MS was used for the rapid dereplication of a wide

range of phytochemicals in leave and roots extracts of thirty-five P vittata individuals A

total of 359 compounds were successfully identified and semi-quantified based on the selected reaction monitoring conditions of 879 home-made reference substances which was considered as a database These files were used as input data to make the correlation analysis between metabolites content and antibacterial activity of these extracts

2.8 Data analysis

The metabolic profile and intensity of 70 methanolic extracts were ready for statistical analysis The bacterial inhibition percentage of 70 extracts was collected in one file by Microsoft Excel version 2016

Paired t-test (two tailed) was used to determine if there is a significant difference between the effect of root and leave extracts of individual plant on the growth of bacteria

11

All files were converted as comma-separated values (csv) files that were considered as input data for statistical analysis done with RStudio version 3.5.3 (http://www.R-project.org/)

Data was normalized by taking their logarithm base 10 The R package pls and mixOmics

(Rohart et al 2017) was used to fit PLS (Partial Least Squares) regression models (plsr and pls function, respectively), to evaluate their correlation between the secondary metabolites content with antibacterial activity and to generate circular correlation circle plots (plotVar function) Pearson correlation coefficient statistics was also set up as a comparison method

12

III Results:

3.1 Antibacterial effect of P.vittata roots and leave extracts on resistance strain

Stenotrophomonas maltophilia K279a

Figure 1 Microdilution testing of P vittata root and leave extract on the growth curve of S maltophilia strain K279a The growth curve shows S maltophilia growth in the treatment of leave (figure 1A) and roots (figure 1B) extracts at the

concentration of 512 μg/mL in 30hs A: aerial part (leave extracts) S: Subterranean parts (root extracts) Gentamicin curve: Growth curve of S maltophilia K279a with Gentamicin 32 μg/mL treatment (positive control) DMSO curve: bacteria in DMSO 2.5% and TSB (negative control) Strain K279a curve: growth curve of strain in TSB.

antibacterial effects at a moderate extent (fig 1A) Meanwhile, this growth inhibition effect

was shown in more samples of roots, with exact 19 negative activity samples in contrast

with 16 samples to promote the growth rate (fig 1B) (Supplement table S1 for classification

of group P and I) For both the leave and roots treated cultivate wells, it is interesting to note that when the lag phases of strain K279a and inhibiting group (I) share the same period

of 7-8h, the advanced promoting group (P) experiences a very rapid lag phase of just approximately 1.5h Also, the log phase of this group (P) lasts from 2.5-16h (13.5h) with fluctuated trend including a non-linear point at 6h, which is comparable with the log phase

of strain and inhibiting group (I) (8-22h) at approximately 14h In comparison between leave and root extracts, the effect of group (P) of root extracts was more profound with growth rate increasing more than 45% Inversely, the inhibiting effect of leave extracts outweigh the roots extracts Notably, extract sample A21 showed the highest inhibition tendency with a maximum of 45% OD value downward with just 7.5 h of log phase Noted

that S maltophilia strain K279 normally goes to death phase after 24h, the system measured

up to 30h of growth curve and determined the bacterial inhibition at 24h time point (Mahdi 2014) The graph also shows that gentamicin inhibits bacterial growth effectively at tested concentration of 32 μg/mL However, there is a significant difference between bacterial growth curve in DMSO 2.5%/TSB and bacterial growth in TSB alone (blank), which indicates the effect of DMSO to be further discussed

14

Figure 2: The inhibition percentage of P vittata leave and roots extracts on the growth of S maltophilia strain K279a The

bars chart indicates the ratio between S maltophilia K279a growth in leave extracts (2A) and root extracts (2B) of P Vittata and

the strain growth alone in the negative control (DMSO) at the 24h time point A: aerial part (leave extracts) S: Subterranean parts

(root extracts) Gen: %inhibition of S maltophilia K279a with Gentamicin 32 μg/mL treatment (positive control)

DMSO: %inhibition of bacterial growth without any treatment (negative control) Standard deviation of triplicate measurements is

also shown on the graphs

15

From the OD value of bacterial growth, the percentage of bacterial inhibition was calculated

as the ratio of bacterial growth in root and leave extracts to the bacterial growth in negative control (in DMSO 2.5%/TSB) at 24h Significantly, the leave extracts inhibition percentage

reached 46.34% (in sample A21) (fig 2a), 1,5 times exceeded the maximum inhibitory of root extracts, at 27.46% (in the sample S3) (fig 2B) The minimum inhibition percentage

of leave and roots extracts follow the same manner, at 4.04% (A19) and 0.83% (S6), respectively The sample of both leaves and roots of number (A/S) 2, 3, 9, 10, 11, 12, 13,

14, 15, 16, 17, 18, 19, 20, 21 all have inhibitory effect In contrast, the root extracts have higher growth promoting effect than the leave extracts with a marked percentage of 47.32% (- S31) in comparison with 32.98% (-A34) in leave extract The growth promoting samples are (A/S) 4, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35 (supplement table S2) Interestingly, even though derived from the same plant number 11, the root extract shows inhibitory effect (S11) while the leave extract shows growth promoting one (A11) Plant sample 5 (A5 and S5) also experience this phenomena at a milder extent

Table I Paired t-Test statistic result on the difference between leave and root extracts antibacterial activity

The inhibition percentage of extracts from

We conducted Paired t-Test statistics to evaluate the statistical difference of antibacterial

activity between the leave and root extracts of P vittata on resistance strain S maltophilia K279a The percentage of inhibition on bacterial growth is the input data set for t-test

16

performing on two groups (leaves and roots) with the same observation number (sample individual) DataAnalysis TookPak on Excel version 2016 is the Add-in to calculate the p

value, which results in a two tailed p-value > 0.05 (Table I) This result indicates that we

can not conclude a statistically significant difference between the inhibition effect on bacterial growth of leave and root extracts Interestingly, means value reported a negative value in inhibition strength for both roots and leave extracts, at -7.37% and -4.78%, respectively

17

3.2 Correlation between antibacterial activity of 70 P vittata extracts and its

metabolic profile:

Figure 3: Correlation circles of PLS regression between metabolites content (M) and the inhibition percentage of

Stenotophomonas maltophilia K279a growth by leave and roots extracts (Y) The 2 upper figures indicate the correlation of

all metabolites including primary and secondary metabolites (3A) or the secondary metabolites separately (3B) with the antibacterial activity of leave extracts The 2 lower figures indicate the correlation of all metabolites including primary and secondary metabolites (3C) or secondary metabolites separately (3D) with the activity of root extract The footnote shows the color codes for each metabolic group The central cross arrows indicates the quadrant label for the plot representation (following clockwise direction: from quadrant I, II, III to IV)

IV III

18

The correlation circle plots (figure 3) of partial least square regression model are the result

of R packages mixomics (plotVar function) to generate the relationship between 359 metabolites (consists of both primary and secondary metabolites) and the evaluated

inhibition percentage of 35 root and 35 leave extracts of P vittata The relationship of two

variables is presented by the inner product of their vectors, which are simplified by marking just the endpoint of the variable vectors (metabolites) These metabolites are projected inside a 2 components space spanned by a circle of radius 1 (equal to a correlation coefficient 1), a radius 0.5 is also included for better visualization The correlation is positive if the angle between vectors (i.e metabolites and Y point) is sharp, negative if the angle is obtuse and null (having no meaning) if the angle is right, 90° Therefore, following the cross-arrows quadrant separation, metabolites fall into quadrant number II can be predicted as having positive correlation with antibacterial activity (Y) Quadrant number

IV predicts negative relationship with Y (activates bacterial growth) and metabolites in quadrant number I and III have no significant meaning for the prediction Only the metabolites closely located to the circle of radius 1 can be directly interpreted, other metabolites located near the origin should be predicted using different dimensions The stronger adjacence of metabolites on the plane indicates their strong positive correlation to each other

In general, the primary metabolite group makes no significant contribution in the prediction

of antibacterial activity of the hyperaccumulator plant (fig 3A, 3C for both primary and

secondary metabolites of leave and root extracts) Whereas, the orange color coded for flavonoid compounds dominates quadrant II envisions their significant role for antibacterial activity With all of 4 predicting plots, M274 (Betaine) shows the most negative correlation with Y in the quadrant IV by a furthest distance Betaine is an osmoprotectant accumulated

by multiple bacterial families which is well recorded to independently stimulate the bacterial growth (Kawahara et al 1990; Roberts 2005) The negative correlation of betaine

and antibacterial activity (Y) is the most significant for leave extracts (fig 3A)