DNA extraction from low-biomass carbonate rock: An improved method with reduced contamination and the low-biomass contaminant database

Abstract

Caves represent a unique environment in which to study subsurface geomicrobial interactions and processes. One of the primary techniques used to study such geologic samples is molecular phylogenetic analysis, but this technique is hampered by low microbial biomass and calcium in the host rock, often leading to poor and irreproducible DNA extraction. We describe an improved protocol to recover extremely low amounts of DNA from calcium-rich geologic samples. This protocol relies on the use of the synthetic DNA molecule poly-dIdC, to act both as blocking agent and carrier molecule to increase the yield of DNA, and dialysis to remove calcium inhibitors of PCR amplification. Further, we demonstrate that many traditionally used laboratory substrates contain microbial DNA that can be amplified through the polymerase chain reaction (PCR) and contaminate molecular phylogenetic profiles. While the number of potential contaminants can be minimized, it cannot be eliminated from extraction techniques. We have therefore established the low-biomass contaminant (LBC) database, which contains the 16S rRNA gene sequences of species that have been identified as common laboratory contaminants. These identified contaminants provide a reference database to allow investigators to critically evaluate certain species identified within their phylogenetic profile when examining such low-biomass environments.

Introduction

It was recently proposed that a substantial portion of the Earth's biosphere is microbial and subterranean (Gold, 1999, Whitman et al., 1998). The discovery of such a significant geomicrobial biosphere should hardly be surprising, given that we have known for some time that microbial interactions have shaped the global environment in which we live today (Ben-Ari, 2002, Schopf, 1983). This hypothesis has been supported by studies that reveal complex subsurface microbial ecosystems, leading to a greater interest in studying geomicrobial interactions in such environments (Banfield and Nealson, 1997, Colwell et al., 1997, Gold, 1999, Stevens, 1997). These studies have generally not relied on microbial cultivation, but have been carried out using environmental DNA extraction. Such DNA can reveal the complexity of microbial communities living in these environments through PCR amplification of the 16S ribosomal RNA (16S rRNA) gene sequence and molecular phylogenetic techniques (Cunningham et al., 1995, Jones, 2001, Northup and Lavoie, 2001, Pace, 1997, Sarbu et al., 1996). Extracting DNA from chemically complex geological samples is technically challenging and as a result these investigations have tended to concentrate on microbial communities displaying significant growth (Angert et al., 1998, Barton and Luiszer, 2005, Holmes et al., 2001, Sarbu et al., 1996). Nonetheless, there is an emerging number of studies that suggest oligotrophic caves still contain significant and diverse microbial communities (Barton et al., 2004, Chelius and Moore, 2004, Schabereiter-Gurtner et al., 2002).

Among the factors that limit the ability to extract DNA from cave environments are the low numbers of bacterial cells generally associated with oligotrophic cave communities (∼ 10⁶ cells cm^− 3) and the chemical nature of the rock (Barton et al., 2004). The majority of caves form in calcium carbonate (CaCO₃), otherwise known as limestone rock (Klimchouk et al., 2000), which binds DNA with a high affinity. This is due to the strong chelating affinity of calcium ions for the phosphate-backbone of the DNA molecule (Martinson, 1973). Thus, the rock powder binds DNA molecules during extraction, carrying them into the insoluble fraction during DNA purification. This is particularly challenging in the context of oligotrophic environments: due to the carrier effect of DNA itself, as the total DNA concentration drops, so does the ability to obtain an extractable amount of DNA. This presents a significant obstacle for studying microbial communities in extremely low biomass, calcium-rich cave environments (Barton et al., 2004).

To overcome the problem of DNA sequestration in calcium-rich coral communities, Guthrie et al. (2000) developed an elegant extraction technique, using the same techniques to prevent non-specific DNA binding that have historically been used to prevent such problems in Southern blotting experiments (Maniatis et al., 1989). Namely, these investigators used blocking agents such as Denhardt's solution and milk powder to coat the reactive surfaces of CaCO₃ and block DNA binding, increasing the yield of DNA (Guthrie et al., 2000). This has proved to be a particularly useful technique in the context of coral reef studies, where microbial communities approach 10⁸–10⁹ bacterial cells per gram of sample. In cave samples, where microbial communities are significantly smaller, such techniques increase the risk of introducing bacterial contaminants that will artificially distort the structure of the community being examined (Barton et al., 2004, Chelius and Moore, 2004, Tanner et al., 1998).

In this paper, we describe a modification upon the Guthrie et al. method, which takes into account the low microbial biomass and chemically complex nature of cave and other samples extracted from sedimentary rock. We have also developed mechanisms to increase the yield of DNA appropriate for PCR amplification, while eliminating many potential sources of 16S rRNA gene contamination. Finally, we have established a low-biomass contaminant (LBC) database. This database includes samples from negative extraction controls and provides a reference source for other researchers studying similar low-biomass environments.