Optimizing ultracentrifugation conditions for DNA-based stable isotope probing (DNA-SIP)

Highlights

• Optimal centrifugation conditions for DNA-SIP of soil sample were explored.

• Centrifugation for 48 h was enough for DNA equilibrium with the Vti 65.2 rotor.

• The optimal initial buoyant density for DNA-SIP is 1.71 g cm⁻³.

• Low centrifuge speed was better for widespread of DNA in gradients.

Abstract

DNA-SIP (DNA-based stable isotope probing) is increasingly being employed in soil microbial ecology to identify those microbes assimilating the ¹³C/¹⁵N labelled substrate. Isopycnic gradient centrifugation is the primary experimental process for conducting DNA-SIP. However, diverse centrifugal conditions have been used in various recent studies. In order to get the optimum conditions of centrifugation for DNA-SIP, centrifugation time (36, 42, 48, 60 h), speed (45,000, 55,000 rpm) and the initial buoyant density (1.69, 1.71, 1.725 g ml⁻¹), as were used extensively in related studies, were tested in this experiment with the Vti 65.2 rotor. DNA with either ¹³C-labelling or unlabelled was extracted from a paddy soil pre-incubated with either ¹³C-labelled or natural abundance glucose. After ultracentrifugation, the gene abundance of bacterial 16S rRNA, fungal 18S rRNA, bacterial and archaeal amoA within the fractioned DNA was detected. The results showed that centrifugation for 48 h was enough for the DNA to reach stabilization in the CsCl solution. The initial density of the mixed solution was best adjusted to 1.71 g ml⁻¹ to ensure that most of the genes were concentrated on the middle fractions of the density gradient. Increasing the centrifugation speed would increase the density gradient of fractions; therefore, 45,000 rpm (184,000 g) was recommended so as to obtain the more widespread pattern of DNA in the centrifugal tube. We hope these findings will assist future researchers to conduct optimum ultracentrifugation for DNA-SIP.

Introduction

Since the DNA-based stable isotope probing (DNA-SIP) technique was first used in microbial ecological studies (Radajewski et al., 2000), it has been employed universally as a tool to link the identification of microorganisms with specific environmental processes in recent years (Coyotzi et al., 2016; Da Costa et al., 2018; Neufeld et al., 2007b). Nowadays, a series of ¹³C (seldom ¹⁵N/¹⁸O)-labelled organic and inorganic compounds are chosen to find those microorganisms that prefer to utilize these materials, such as carbon sources (glucose, methanol, butyrate, xylose, cellulose), organic contaminants (isoprene, PCP, PAHs), plant residues, root exudates, CO₂, CH₄ or urea (Esson et al., 2016; Haichar et al., 2012; Li et al., 2015a; Li et al., 2015b; Long et al., 2018; Nicholson et al., 2009; Pepe-Ranney et al., 2016; Radajewski et al., 2000; Song et al., 2016; Zhao et al., 2015).

Except for a small number of pure culture experiments, most studies have been conducted in complicated ecosystems, such as soil, sediments, waters and bioreactors (Jameson et al., 2019; Li et al., 2019; Niu et al., 2019; Sultana et al., 2019; Xu et al., 2019). Generally, after the sample has been incubated with these substrates for several hours to weeks (Jiang et al., 2015; Zhang et al., 2015), DNA of the soil microbes is extracted and mixed with CsCl and gradient buffer. Ideally, the DNA of microorganisms assimilating the ¹³C-labelled substrates would show up in relatively heavier density fractions after isopycnic ultracentrifugation compared to corresponding natural ¹³C abundance substrates. Subsequently, the labelled DNA is picked out and analyzed by clone library or high throughput sequencing to identify the specific microbial population metabolizing the related carbon source.

The key step in these procedures is to identify the labelled fractions after ultracentrifugation. In the first report of DNA-SIP for ecological studies, two-step ultracentrifugation, combined with ethidium bromide (EtBr) staining, was adopted to distinguish the DNA in different fractions visibly (Radajewski et al., 2000). Considering the accuracy and sensitivity, Lueders et al. (2004) selected fluorometry and real-time PCR to in order to detect the ‘heavy’ nucleic acids in a short time. Although the experiment was based on pure cultures, this approach was also widely used in soil microbial studies (Gkarmiri et al., 2017; Hannula et al., 2017; Sun et al., 2018). Due to the complexity of soil microbial community structure and function, it is not so easy to find the labelled fractions. It is significant that it was observed that almost 8 gradient fractions shifted after ultracentrifugation for pure cultures (Lueders et al., 2004) but only 1––4 for environmental samples (Hu and Lu, 2015). Therefore, DNA-SIP for environmental samples should be implemented more cautiously than for pure culture. Otherwise, there was a puzzling phenomenon that labelled 16S rRNA gene or DNA abundance could be identified after labelling root exudates for 14 days or 8 days in a pulse experiment, or even labelling with CO₂ for 28 days (Beulig et al., 2015; Da Costa et al., 2018; Gschwendtner et al., 2016), while it was hard to separate the curves of qPCR between the ¹³C and ¹²C treatments after continuously labelling root exudates for 21 days with different soils and plants in our laboratory (Wang et al., 2019). Although the condition of soils and plants were clearly important factors, the labelling amount of soil microbes by continuous labelling was theoretically higher than pulse labelling of plant and ¹³CO₂ of bulk soil. In this case, the failure in detecting the presence of labelled DNA in Wang et al. (2019) was unlikely due to the insufficient labelling amount of DNAs but maybe caused by nonoptimal ultracentrifugation procedures.

As known, the separation degree of labelled nucleic acid fragments depends on the rotor geometry, isotopic label (¹³C/¹⁵N/¹⁸O), and type of nucleic acid (RNA/DNA). Here we selected the Vti 65.2 rotor to study the impact of ultracentrifugation conditions on isopyncnic gradients. Although the protocol for DNA-SIP was initially published (Neufeld et al., 2007a), there has been much divergence in ultracentrifugation procedures existing in different labs. For example, the density of the blended solution (containing DNA, CsCl, and gradient buffer) before ultracentrifugation was adjusted to 1.69 to–1.77 g ml⁻¹, the time of ultracentrifugation ranged from 24 to 66 h, and the speeds of ultracentrifugation were set from 43,300 to 56,200 rpm (about from 140,000 to 228,000 g). In order to examine these factors, soil was incubated with ¹³C- glucose to ensure that the DNA of the soil was strongly labelled with ¹³C, then different initial densities of solution, and different speeds and times of ultracentrifugation were designed to compare the separation effect of universal genes (bacterial 16S rRNA gene and fungal 18S rRNA gene, existing in diverse microbes) and functional genes (bacterial and archaeal amoA gene, only existing in some specific microbes) in this experiment. The aim was to provide suggestions for optimum ultracentrifugation conditions for ecological studies employing DNA-SIP.