Isolating Bacterial RNA for Microarray Analysis
The rapid increase in the number of bacterial genomes sequenced over the past several years has opened up a new era in whole-genome expression analysis using DNA microarray technology. In recent years, this tool has been used to explore transcriptional profiles for a variety of bacteria leading to a greater understanding of microbial metabolism [1]. DNA microarray technology has also led to a better understanding of microbial pathogenesis, physiology, epidemiology, ecology, fermentation, and pathway engineering.

With the emerging importance of bacterial microarray gene expression analysis, it is vital that bacterial RNA isolated for array analysis be of the highest possible quality. Key characteristics unique to bacteria known to significantly impact RNA quality, and dramatically influence the quality and validity of array analysis, include:

  1. Growth phase at harvest and harvest conditions
  2. Rapid turnover of bacterial mRNA
  3. Bacterial mRNAs' lack of a poly(A) tail, thus excluding the use of classical enrichment methods (e.g. oligo(dT) enrichment)

The last of these factors was the basis for developing Ambion's MICROBExpress™ Kit. MICROB Express dramatically increases the sensitivity of array analysis by removing >95% of 16S and 23S rRNA from bacterial total RNA, leaving a highly enriched population of bacterial mRNA. The enriched mRNA serves as a superior template for synthesizing the labeled cDNA used in whole genome array analysis.

Pre-Isolation Factors That Affect RNA Quality

There are several key factors to consider before beginning any bacterial RNA isolation procedure. The first is when to harvest the bacteria to
maximize yield and quality of intact RNA. The highest quality RNA is isolated from cells in the logarithmic phase of growth. RNA isolated from bacteria in the stationary phase typically exhibits increased degradation. For this reason, it is highly recommended that RNA not be harvested from cells that have grown into stationary phase unless a specific experiment requires such isolation.

A second factor to consider is the method used to harvest the bacteria. It is known that bacterial mRNAs exhibit a wide range of stabilities. For example, approximately 80% of all mRNAs in E. coli are known to have half-lives of between 3 and 8 minutes [2]. For this reason, it is important to collect cells in a manner that minimizes or eliminates the impact of harvesting conditions on both gene expression profiles and/or RNA quality. With this in mind, it is preferable to process small volumes of bacterial cultures (1-2 ml) and to use a brief centrifugation step (~1 min.) to pellet the cells. The cell pellets can then be processed with one of the following treatments:

  • Immediate cell lysis and RNA purification using RiboPure™-Bacteria Kit
  • Rapid freezing in liquid nitrogen (a freeze-thaw treatment may aid lysis of some bacteria)
  • Resuspension of cells in RNAlater®.

These simple steps are of vital importance for obtaining the highest quality RNA required in array analysis.

Isolating Bacterial Total RNA: RiboPure™-Bacteria

Traditional methods for isolation of total RNA from bacteria require the use of hot phenol or include other harsh techniques, some of which need specialized equipment. For Gram-positive organisms, pretreatments with enzymes to weaken the cell wall are often used prior to isolation of RNA. However, it should be kept in mind that these enzymatic pretreatments may alter gene expression patterns on a global scale and could compromise array data.

The RiboPure-Bacteria RNA Isolation Kit uses no enzymatic pretreatments and thus will not alter gene expression profiles. RiboPure-Bacteria combines an efficient glass bead and phenol-based RNAWIZ™-mediated disruption step followed by a glass fiber filter-based RNA purification step for high yields of exceptionally pure RNA.

Removing Contaminating Genomic DNA: TURBO DNA-free™

Once the RNA is purified, it should be treated with DNase to remove any contaminating genomic DNA. This is especially important for array analysis and/or qRT-PCR. Contaminating genomic DNA will be labeled during the cDNA synthesis/labeling reaction and/or lead to non-specific RT-PCR products that could interfere with quantitative results. TURBO DNA-free is ideal for this application.

Enriching for Bacterial mRNA: MICROBExpress™

For decades mRNA has been isolated from eukaryotic sources using oligo(dT) selection. Bacterial mRNAs, however, lack the relatively stable poly(A) tails found on eukaryotic messages and thus cannot be enriched using affinity selection with oligo(dT). Ambion's MICROBExpress Kit offers enhanced signal sensitivity in gene expression analysis across all array platforms tested (including glass, nylon and GeneChip®) by its ability to remove >95% of bacterial rRNA from total bacterial RNA preparations. The depletion of rRNA leaves behind a highly enriched bacterial mRNA population (See Figures 1 and Figure 2). Additionally, MICROBExpress leads to an increase in the total number of genes detected (i.e. increase in % present calls) and an improvement in overall signal to noise on GeneChip E. coli Antisense Genome Arrays (Figure 3).


Figure 1. Agilent® Bioanalyzer Image of E. coli mRNA Purified Using MICROBExpress™. 16S and 23S rRNAs were selectively removed from five different 10 µg samples of E. coli total RNA and the remaining enriched mRNA fraction (~70 ng each) were run in Lanes 2-6. A sixth 10 µg sample of the same total RNA prep was subjected to a mock MICROB Express procedure without using the kit's rRNA capture oligo mix. Approximately 0.7 µg micrograms of the mock reaction was loaded in Lane 7. As seen in the gel image the rRNA bands are virtually undetectable in the mRNA obtained with the MICROB Express procedure versus the control lane.



Figure 2. Agilent® Bioanalyzer Electropherograms of mRNA from Figure 1. As seen in the (top) electropherogram, greater than 95% of the bacterial 16S and 23S rRNA bands have been removed, leading to a selective enrichment of mRNA species as compared to the (bottom) mock MICROB Express™ procedure. (Relative fluorescence units have been rescaled between the two electropherograms to allow a closer examination of the MICROB Express enriched RNA).



Figure 3. Affymetrix® GeneChip® E. coli Antisense Genome Array Analysis. Illustrated above are Log Plots (Signal vs. Signal) for technical replicate GeneChips (i.e. same RNA, different probe labeling used for replicates). Panel A compares replicate arrays using 10 µg RiboPure™-Bacteria isolated E. coli total RNA with the standard Affymetrix labeling and hybridization protocol. Panel B compares replicate arrays utilizing the mRNA (0.69 µg) obtained with the MICROB Express™ Kit from 10 µg of RiboPure-Bacteria isolated total RNA; this sample was labeled and hybridized using the same protocol as in Panel A. The total RNA (average between duplicate GeneChips) gave 73% Present Calls (genes above background) versus MICROB Express isolated mRNA which gave 80% Present Calls, thus demonstrating an increase in sensitivity. In addition, the number of Concordant Calls for duplicate GeneChips increased with MICROB Express mRNA (94%) compared with total RNA (92%). Concordant Calls are those genes called identically on both arrays (Present/Present, Marginal/Marginal or Absent/Absent), and is a direct measure of reproducibility.


*Correlation coefficients were >0.99 for both replicate comparisons.
GeneChip® is a registered trademark of Affymetrix, Inc.