Loading...

Summer Research Fellowship Programme of India's Science Academies

Assessment of NF-κB expression in monocytic cell line

Yashasvi Rao

B.Sc Biotechnology, Dr. Ambedkar College, Nagpur 440012

Dr. Prasenjit Guchhait

Professor, Regional Centre for Biotechnology, Faridabad, Haryana 121001

Abstract

Nuclear Factor kappa-light-chain-enhancer of activated B cells (NF-kB) is a transcription factor that regulates the expression of hundreds of other genes. NF-kB is a dimeric protein complex, made by the combination of five constituent proteins which are p50 (NF-kB1), p52 (NF-kB2), p65 (RelA), RelB and c-Rel. These proteins can bind with each other to form hetero- or homo-dimers and then bind to a specific sequence of DNA in a gene to regulate its expression. The type of regulation depends on the combination of proteins in NF-kB. The homo-dimer p65(RELA)-p65(RELA) regulates by activating the translation of its target gene. NF-kB binds to different genes that code for mediators involved in inflammatory and immune response. Hence NF-kB plays a crucial role in the formation of many mediators that participate in inflammatory and immune response. Some of these mediators are cytokines, chemokines, etc. The NF-kB complex when in its inactive state, is present in the cell cytoplasm and is bound by inhibitors (NF-kappa-B inhibitors i.e I-kappa-B). The inhibitor dissociates from the complex after a phosphorylation step and the complex is then activated. In its activated state it enters the cell’s nucleus and binds to the DNA molecule specifically at the kappa site present on it. NF-kB is a ubiquitous protein hence it can be found in many cells. Monocytes are one of the important constituents of innate immunity. One of their functions is synthesis of cytokines and chemokines, therefore in monocytes, NF-kB is found in abundance. So in our project, we will be using monocytic cell lines to estimate the gene expression of RELA. The
expression will be measured for activated as well as inactive monocytes and the
technique used will be real time polymerase chain reaction (RT-PCR).

Keywords: RELA, monocytes, RT-PCR, gene expression, innate immunity, transcription factor

Abbreviations

Abbreviations
 LPSLipopolysaccharide
 RT-PCRReal Time Polyperase Chain Reaction

INTRODUCTION

Background

The immune system has been classified into two classes: Innate immune system and adaptive immune system. As the name suggests, innate immunity is something an individual has inbuilt since birth. On the other hand acquired immunity develops in an individual through the course of time by encountering several foreign antigens. The components of these two immunities are different. The innate immunity has four modes of defense which are physiological (temperature, pH etc.), anatomical (skin, mucous membranes), inflammatory and phagocytic (monocytes, macrophages, neutrophils). These barriers are not specific in action. They defend the host from a category of pathogens that show a frequently encountered class of cellular or molecular components. The adaptive immunity is more specific in its action. The components of adaptive immunity are various cells and organs of immunity.

Role of monocytes in innate immunity

Monocytes are phagocytic in nature and are derived from myeloid progenitor cells during hematopoiesis. Monocytes keep circulating in the blood before activation. Monocytes have receptors on their membrane surface. these receptors are pattern recognition receptors (PRR) or toll like receptors (TLR). When various microbial components, or foreign substances enter the host, they get bound to these receptors present on the monocyte surface and activate them. Upon activation via stimuli such as LPS, TNF-α etc. monocytes differentiate into macrophages which then migrate to the specific tissue and become stationary.

Statement of the Problems

The phagocytic activity of cultured monocytes is enhanced upon stimulation of purified bacterial LPS(1). LPS stimulation of human monocytes activates several intracellular signaling pathways that include the I-kappa-B kinase (IKK)–NF-κB pathway and this signaling pathway in turn activates transcription factor NF-κB (p50/p65) which coordinates the induction of many genes encoding inflammatory mediators(2). The majority of NF-κB complexes are made of p50-p65 dimer. The subinit p50 is coded by gene NF-κB1 while p65 is coded by the gene RELA.

NF-κB

Nuclear Factor kappa-light-chain-enhancer-in-B cells or NF-κB is an omnipresent transcription factor that regulates the expression of a wide number of genes taking part in the synthesis of several immune and inflammatory mediators. NF-κB induces the expression of various pro-inflammatory genes, including those encoding cytokines and chemokines, and also participates in inflammasome regulation (3). NF-κB is a dimeric protein complex, found as a combination of its subunits p50/NF-κB1, p52/NF-κB2, p65/RELA, RelB or c-Rel. These can either have two same subunits (homo-dimer) or two different subunits (hetero-dimer), and depending on the combination, the complex can work either as a transcriptional activator or repressor, but in most cases it acts as an activator. The majority of NF-κB complexes were seen to be a combination of p50-p65 dimer. The NF-κB complex is present inactivated by its inhibitors in the cell cytoplasm before activation. The complex is bound by inhibitors of IκB family.

Objective of the Research

To check whether monocyte stimulation would result in differential expression of NF-κB

LITERATURE REVIEW

Information

nf-kb pathway.jpg
    The diagram above illustrates the canonical pathway of NF-κB proteasomal degradation of IκBα.

    Based on the type of activation, there are two pathways through which NF-κB can be activated (4). The first is canonical pathway which is less specific and has a wide range of stimulus recognition while the non-canonical pathway or the alternative pathway is relatively more specific.

    NF-κB complexes are found in the cytoplasm in an inactive state, bound by their inhibitors. These inhibitors belong to a family of NF-kappa-B-inhibitors (IκB). The canonical pathway activates the NF-κB complex by phosphorylation of these inhibitors so that they dissociate from the complex. This dissociation is accomplished with the help of another family of enzymes called I-kappa-B Kinases (IκK). This enzyme (IκK) itself is made of three subunits: IκKα, IκKβ, and NEMO (NF-κB essential Modulator) or IκKγ (5). When a cell is exposed to any kind of stimulus such as LPS, Pattern Recognition Receptors (PRR), etc IκK gets activated. It specifically phosphorylates the inhibitor subunit IκBα at two N-terminal serine residues, which leads to its proteasomal degradation, rendering the NF-κB complex active (6)(7). The activated NF-κB complex then transiently migrates to the nucleus and binds to kappa B sites present on the DNA to regulate the transcription of its respective target gene. The non-canonical pathway on the other hand, responds to specific stimuli and doesn’t involve IκBα degradation. It modifies the precursor of NF-κB2 viz. p100. This p100 is phosphorylated to remove its IκB like structure at the C-terminal to give mature NF-κB/p52 molecule. It then migrates to the nucleus and binds to the kappa-B sites on DNA.

    Monocytes are important cells of immune system that play a crucial role in defense mechanism by their phagocytic and antigen presenting properties, therefore they would have a greater amount of NF-κB present in them. Hence we have opted to perform our experiment on Monocytic cell line THP-1.

    thp-1 cells.jpg
    • 1
    • 2
    Microscopic image of THP-1 cells

    METHODOLOGY

    Concepts

    • DAY 1: Seeding and culture of THP-1 cells in RPMI medium for 24 hours
    • DAY 2: Group A- LPS treated THP-1 cells. Kept under stimulation for 8 hours at 5% CO2, 21% O2, 805 Humidity. Group B:- untreated THP-1 cells
    • RNA isolation: - Cell collection and RNA isolation by Trizol Method and quality check of isolated RNA by 0.5% bleach gel electrophoresis
    • cDNA :- Complementary DNA preparation from isolated RNA ·       
    • RT-PCR: - Quantitative PCR of RELA mRNA transcript and Calculation of Gene expression

    Methods

    Culture of THP-1 Cells

    Day 1: Seeding and culture of Monocytic cell line THP-1

    THP-1 is a well-established human leukemic cell line derived from one-year-old male patient suffering from Acute Monocytic Leukemia (AML) (8). These cells show quite similar properties as the normal monocytes hence they are a very efficient alternative for in vitro study of functions and mode of action of monocytes under infection or disease. The cells were grown in RPMI medium, at 5% CO2, 21% O2, 80% humidity with FBS (fetal bovine serum) at 37 degrees Celsius for 24 hours. The cells were allowed to reach a population of about 0.5 million per well before collection.

    Day 2: We made two groups of the cultured cells. The first group was treated with lipopolysaccharide (LPS). LPS is a component of gram negative bacterial cell wall, hence it would help in eliciting an immune response from the monocytes, which are cells of immune system. The first group was kept under stimulation of 1 μg/mL LPS for 8 hours under normal cell culture conditions. After 8 hours, the cells were collected. The plate was placed on a rocker to remove any cells that may be attached to the plate. The cells were collected from the wells using a pipette. The cells were mixed with the pipette to again remove any attached cells as LPS activated cells are slightly stickier. The cells along with their medium were transferred to new 1.5 mL microfuge tubes.

    RNA isolation by using Trizol

    The cells were then centrifuged at a speed of 6000 rpm for 5min to give a white colored pellet. The supernatant was discarded and 1mL Trizol was added to each tube. Trizol is a monophasic mixture of acidic phenol and guanidium thiocyanate(9). Guanidium thiocyanate is a chaotropic agent that can denature the membrane proteins to disrupt the cell membrane and expel its contents. As the cells bursts open, DNA, RNA, proteins and other constituents come out in the surrounding solution. Since phenol is only partially soluble in water, it forms a separate layer beneath it. At pH 7-8 DNA has negative charge because of the phosphate group. But at acidic pH of 4 - 4.5 the negative charge on DNA is neutralized and it dissolves in phenol. Although RNA remains in the aqueous phase as the pKa value of the functional groups present in it is slightly higher and do not get neutralized at this pH. Hence RNA and DNA get separated from each other too. After this, 0.2 mL chloroform per 1 mL Trizol is added to the tubes. Phenol is partially soluble in water, so it also retains some amount of water with dissolved RNA. The chloroform dissolves phenol but is denser than water, helping in sharper separation of the phenol and water phases. The tubes are then incubated for 2-3 minutes and then centrifuged for 15 minutes at 12000 g at 4°C. Now three distinct layers can be visualized in all the tubes. The upper aqueous and colorless phase has RNA, The interphase has DNA and the lower pink organic layer has proteins and DNA. The upper aqueous layer is transferred using a pipette into a new microfuge tube of 1.5 mL. This step must be carried out with extra caution as contamination of RNA with DNA from interphase is most probable. After the transfer to new tubes, 0.5 mL isopropanol is added to the aqueous phase and incubated for 10 minutes. Isopropanol helps in the precipitation of RNA as it doesn’t dissolve in it. The tubes are centrifuged again at least at 13.2 x 1000 rpm for 30 minutes at 4°C. RNA could be seen precipitated as whitish gel like pellet. The supernatant is discarded using a pipette and the pellet is re-suspended in 1mL of 75% ethanol. Ethanol washes the pellet by dissolving all the water soluble substances that may be present in the pellet. The tubes are put on a vortex briefly. Then again they are centrifuged at 7500 g for 5 minutes at 4°C. The pellet now obtained has only RNA as the impurities get dissolved in the ethanol. The supernatant is removed and the pellet is air dried for a while without letting it dry completely. The pellet is re-dissolved in 20-50 μL of nuclease free water.

    Measurement of concentration and purity of isolated RNA

    The concentration of RNA in the samples was measured using Nanodrop equipment that measures the absorbance of the sample at 260nm and 280 nm to estimate the concentration. The ratio of absorbance at 260 and that at 280 nm indicates the purity of the sample. Double stranded DNA absorbs at 280 nm while RNA, ssDNA, and free nucleotides absorb at 260 nm. Hence to ensure a pure RNA sample without contamination of proteins and DNA, the value of 260/280 should be in between 1.8 to 2. A value below 1.8 indicates protein contamination. 

    Quality check of RNA by 0.5% bleach gel electrophoresis

    Bleach gel electrophoresis is done for RNA to check whether it is intact. This is necessary due to the omnipresence of RNase enzyme (10). Using 0.5 to 1% bleach in agarose gel electrophoresis inhibits the RNase that might be present in the gel or sample. Bleach denatures the proteins but it does not affect the nucleic acid present in the sample. The gel concentration used was 2% agarose in 1X TAE buffer, as the RNA is small molecule. ETBR was used as the RNA binding dye at a concentration of 3μL/100mL.

    RNA gel.png
      Result of bleach gel electrophoresis of RNA sample. Two clear and distinct bands for 28S and 18S rRNA can be seen for most of the samples, indicating that RNA in sample is intact.

      cDNA synthesis

      Complementary DNA or cDNA is prepared from mRNA by the action of RNA dependent DNA polymerase (Reverse Transcriptase). This process is called Reverse Transcription because unlike the typical transcription where RNA is formed from DNA, here DNA is formed from RNA. The length of cDNA would be shorter than the genomic DNA as it would not have only the exon (non-coding) regions on it. In this step we have used mixture of oligo(dT) primers and random hexamer primers that bind to RNA molecule at random sites and help in providing a starting point for cDNA formation. In our experiment, only one cycle is carried out in the thermal cycler, therefore only one copy of cDNA was formed. The quality of cDNA depends on the integrity of the messenger RNA and the fidelity with which it can be reverse trancribed. RNA cannot be cloned directly, in a reaction catalyzed by reverse transcriptase, the RNA together with a suitable primer and a supply of dNTPs, must be converted to a double-stranded molecule (11). We have used a BIO-RAD kit for the cDNA synthesis. The kit contents include a 5X iScript reaction mix consisting of the mixture of oligo(dT) primers and random hexamer primers, dNTPs, etc and Reverse Transcriptase RNase H+ obtained from Moloney Murine Leukaemis virus (MMLV). The protocol used was as follows:

      Cdna cycle.png
        The various steps involved in the cDNA preparation at different temperatures for various time periods.

        Real Time Polymerase Chain Reaction (RT-PCR)

        RT-PCR is a quantitative method to estimate the gene expression. In this technique we use the previously prepared cDNA and amplify it for 40 cycles. Even one copy of a specific sequence can be amplified and detected in PCR. The PCR reaction generates copies of a DNA template exponentially. This results in a quantitative relationship between the amount of starting target sequence and amount of PCR product accumulated at any particular cycle(12). There are three main steps in one cycle of any PCR:  

        1.     Denaturation: This step involves the separation of the dsDNA into ssDNA by the breaking of hydrogen bonds between base pairs of opposite strands. Since this needs breaking of bonds, the temperature at which denaturation occurs is high. By default it is set at 94˚C.

        2.     Annealing: This step consists the binding of the primers to ssDNA so as to provide a binding site for DNA polymerase. The annealing temperature depends on the melting temperature of the primers. In general this step takes place at 50˚C.

        3.     Elongation: In this step there is elongation of the new strand by adding of the dNTPs from surrounding mixture. This reaction is catalyzed by the enzyme DNA polymerase. The elongation step is usually carried out at 72˚C.

        PCR cyclr.png
          Graphical representation of the temperature and time period for steps involved in PCR. 

          Primer design

          In our experiment we are focusing on the expression of RELA which is our target gene, and as housekeeping gene we have used β-Tubulin. So we have selected the primers specific to these genes. The primers used in RT-PCR are sequence specific, hence it is necessary to carefully design the primers in order to get only the desired gene amplification and no non specific amplification. There are a few things to keep in mind before designing the primers for RT-PCR:

          1.     Primer Length – for RT-PCR the preferred primer length is from 18 to 28 bp.

          2.     Tm – Melting temperature of the primers should be within 5˚C of each other.

          3.     Product Length: The desired product length of RT-PCR is in the range of 60-200 bp. For our experiment the product length was 135 bp long.

          4.     Self-coiling: The primer set showing the least coiling is selected.

          5.     GC%: The GC content up to 50% is considered.

          Target Gene Accession number Forward Primer Reverse Primer Product length
          RELA NM_021975.3 TGAACCGAAACTCTGGCAGCTG CATCAGCTTGCGAAAAGGAGCC 135
          TUBB NM_178014.2 CTGGACCGCATCTCTGTGTACT GCCAAAAGGACCTGAGCGAACA 117

          For selecting the primers we first obtained the accession number for that gene with the help of NCBI website. After receiving the accession number, we did a primer blast by mentioning the desirable product length which in this case was 60-200 bp. The website shows a number of primer sets based on the criteria mentioned. With the help of GeneRunner software we checked for GC content, primer dimer formation, loops and self-coiling in these sets. All the sets show some extent of self-coiling but we have selected the set with least self-coiling.

          Quantification of RELA mRNA transcript

          In the normal PCR technique, one puts the sample for amplification and estimates the concentration of DNA at the end by measuring absorbance at 280 nm and then running agarose gel electrophoresis. But in the quantitative or real time PCR one can measure the concentration or copy number of amplicon after completion of every cycle. In this technique the target gene is measured by detecting the fluorescence emitted by it. The dye used in RT-PCR is usually SYBR Green. This dye binds non-specifically to the dsDNA. Therefore, any other non-specific amplification can also be detected by the help of SYBR Green. SYBR Green intercalates with the dsDNA by binding to the minor grooves.

          Reagents Volume
          Primer Mix 2 μL
          Nuclease free water 3.5 μL
          cDNA 4 μL
          SYBR Green 2.5 μL
          Total volume 10±2 μL

          Amplification plot

          This is plotted between the cycle number on the x-axis and relative fluorescence on the y-axis. The graph usually shows noise in the initial cycles. The signal level at which there is almost no change in relative fluorescence is called the baseline. After about 15 cycles the curve shows steady increase. The signal level from which the there is an exponential increase in the fluorescence is called the Threshold. The cycle number at which the fluorescence crosses threshold is called the threshold cycle or C­t. This Ct value is used for further calculations. Ct value is inversely proportional to the amount of target gene present in the sample. Therefore, more the amount of target gene in the sample, lesser the value of Ct, it means that the fluorescence will cross the threshold earlier for it.

          Amplification plot.png
          • 1
          • 2
          • 3
          • 4
          Amplification plot where x-axis shows the cycle number and y-axis has the relative fluorescence.

          Melting curve plot

          This plot uses the difference in the values of melting temperatures for different DNA fragments to detect the presence of non-specific amplicon. The melting temperature (Tm) of a DNA molecule is the temperature at which 50% of the dsDNA becomes single stranded due breaking of the hydrogen bonds between the base pairs. This Tm value of any DNA fragment depends on the number of nucleotides present on it and its GC content. More the number of nucleotides in a DNA fragment higher the Tm value. Similarly the more the GC content in it, higher the Tm value as more hydrogen bonds need to be broken. Since different fragments will have different lengths and GC content, they will also have different Tm ­value. This phenomenon is used to detect the presence of any non-specific amplification taking place along with the target amplification. Sometimes there may be a contamination of genomic DNA along with the extracted RNA in the sample. This genomic DNA may also get amplified along with cDNA making the process inefficient. In some cases there is formation of primer dimers, which would cause an increase in the Ct. Presence of such contaminants can be detected by using a melt curve plot. For example, primer dimers being shorter in length than the target gene will have a lower Tm. The SYBR Green dye binds to dsDNA, so when the temperature crosses the value of Tm, there is a sharp decrease in the fluorescence. Hence a peak is seen at the Tm for that DNA fragment.

          Melt curve plot.png
          • 1
          • 2
          Melt curve plot for primer dimer in the no-template control and of the target gene. The melt curve plot for the no-template control (NTC) is an overlay on the amplification plot of target gene to show the difference in the positions of their respective peaks. The green curve has a peak at 80˚C for primer dimer in NTC. The blue curve shows no other peaks, indicating that the peak obtained is solely for the target gene.

          RESULTS AND DISCUSSION

          The data obtained from RT-PCR was calculated by relative quantification method.

          1.     Method I: In this method, the Ct values are first recorded for all samples, i.e. treated and untreated for both RELA and TUBB. Then the ∆Ct for each sample is calculated by subtracting the Ct value of the TUBB sample from the Ct value RELA sample. This is done for LPS treated and untreated samples. The relative quotient (RQ) is calculated by using the formula 2(-∆Ct). The ratio of RQ values for +LPS and –LPS gives the fold change.

             ∆Ct = difference in the Ct of target gene and housekeeping gene for a sample i.e.

          ∆Ct (+LPS) = (+LPS)Ct(RELA) – (+LPS)Ct(TUBB)

          ∆Ct (-LPS) = (-LPS)Ct(RELA) (-LPS)Ct(TUBB)

            Relative Quotient (RQ) = 2(-∆Ct)

          RQ (+LPS) = 2­­­(-∆Ct)+LPS

          RQ (-LPS) = 2(-∆Ct)-LPS

          Fold change = (RQ +LPS)

          (RQ –LPS)

          2.     Method II: Another method is by calculating the ∆∆Ct value. It is the difference in the ∆Ct values of the treated and untreated samples. The fold change from ∆∆Ct can be calculated as 2(-∆∆Ct).

          ∆∆Ct = ∆Ct (+LPS) - ∆Ct (-LPS)

          Fold change = 2 (-∆∆Ct)

          We have calculated the data using the method I.

          After calculating our data, we have plotted it to get the following graph.

          Bar diagram.png
            The experiment was performed 3 times in duplicates. RT-PCR was done in triplicates. Each bar represents standard error mean. Significance was calculated using paired t-test *<0.05

            CONCLUSION

            RELA translation results in the production of protein p65. p65 partners with other proteins to act as a transcription factor. Our data shows that LPS stimulation does result in the differential expression of RELA in THP-1 cells. The expression of RELA in cells treated with LPS was up-regulated as compared to untreated cells by a factor of 2.38816. This means that upon stimulation with LPS (i.e. bacterial infection), the production of NF-κB is increased in monocytes. So as RELA is up-regulated, it is possible that it orchestrates inflammation due to triggering of downstream molecules by partnering with p50 and others.

            REFERENCES

            1.William E. Bennett & Zanvil A. Cohn: The isolation and selected properties of blood monocytes (Journal of Experimental Medicine, September 1966)

            2.Muasumee. G & Nigel M.:LPS induction of gene expression in human monocytes (Elsevier, February 2001)

            3.Ting Liu, Lingyun Zhang, Donghyun Joo and Shao-Cong Su: NF-κB signaling in inflammation ( Nature, July 2017)

             4.Ebrahim Zandi, David M. Rothwarf, Mireille Delhase, Makio Hayakawa, and Michael Karin: The IkB Kinase Complex (IKK) Contains Two Kinase Subunits, IKKa and IKKb, Necessary for IkB Phosphorylation and NF-kB Activation (Cell Press, October 1997)

             5.Inder M. Verma, Jennifer Stevenson IκB kinase: Beginning, not the end (Proceedings of the National Academy of Sciences of the United States of America, October 1997)

            6.Alaen Esraël: A Role of Phosphorylation and Degradation in control of NF-kB activity (Cell Press issue: Trends in Genetics, June 1995)

            7.Vincent Feng-Sheng Shih, Rachel Tsui, Andrew Caldwell & Alexander Hoffmann: A single NFκB system for both canonical and non-canonical signaling (Nature, November 2010)

             8.Shigeru Tsuchyam, Michiko Yamabey, Yoshiko Yamaguchi, Yasuko Kobayashi, Tasuke Konno and Keiya Tada: Establishment and characterization of human acute monocytic leukemia cell line (THP-1) (International Journal of Cancer, August 1980)

             9.Domenica Simms, Paul E. Cizdziel, Piotr Chomczynski: - TRIZOL ™: A NEW REAGENT FOR OPTIMAL SINGLE-STEP ISOLATION OF RNA (1993)

             10.Patrick S. Aranda, Dollie M. LaJoie, and Cheryl L. Jorcyk: Bleach Gel: A Simple Agarose Gel for Analyzing RNA Quality (January 2012)

             11. Krug, M.S.; Berger, S.L: First-strand cDNA synthesis primed with oligo(dT). (1987)

             12. Arya M, Shergill IS, Williamson M, Gommersall L, Arya N, Patel HR: Basic Principles of real time quantitative PCR. (January 2014)

            ACKNOWLEDGEMENTS

            First of all, I would like to acknowledge my supervisor Dr. Prasenjit Guchhait for giving me the opportunity to work in his lab. I have been trained to not just perform an experiment but also to understand the theory and concept behind it. I would like to thank my seniors and instructors Sulagna Bhattacharya and Nishith Shrimali, who have been busy with their own projects and yet have been guiding me at every step for the last two months. I thank all the members of Disease Biology Lab for their support, Regional Centre for Biotechnology Faridabad for their professional equipment facility and lab space. Lastly, I would like to thank IAS-INSA-NASI for organizing this research fellowship program.

            Source

            • Fig 1: https://www.caymanchem.com/cms/caymanchem/cmsImages/figure1_id153.jpg
            • Fig 2: image source: www.atcc.org or https://www.google.com/search?q=atcc+thp-1+cells&sxsrf=ACYBGNT3_veA7zHziAiJCObZGAOQnZeJQw:1573745874289&source=lnms&tbm=isch&sa=X&ved=0ahUKEwjT3LaehOrlAhWzmOYKHcRIBhwQ_AUIEigB&biw=1366&bih=657#imgrc=hEEgDeaJz-iMLM:
            • Fig 5: image.slidesharecdn.com
            More
            Written, reviewed, revised, proofed and published with