Ancient DNA: Methods and Protocols (3 page)

Base

Multiple extractions

lesions via

miscoding lesions

misincorporations

and amplifi cations;

hydrolysis

Adenine to hypoxanthine

A®G

cloning; UDG (uracil

Cytosine to uracil a

C®T

DNA glycosylase) to

5-methylcytosine to thymine a

C®T

remove uracil

Guanine to xanthine

G®A

Blocking and

Base modifi cations

No amplifi cation;

Special polymerases;

Miscoding

5-OH-5-methylhydantoin

jumping PCR

cloning; multiple

lesions via

(blocking)

amplifi cations

oxidation

5-OH-hydantoin (blocking)

Base

8-oxoguanosine

misincorporation

(miscoding G®T)

Crosslinks

DNA–DNA crosslinks

No amplifi cation

PTB (
N
-phenacy—

via alkylation

lthiazolium bromide) to

DNA–protein crosslinks

cleave crosslinks (but see

(i.e., Maillard products)

( 44 )
)

a Generally, only C®T (or complementary strand G®A) transitions are observed
( 43
) 1 Setting Up an Ancient DNA Laboratory

3

Finally, crosslinks either within or between strands
( 18
) will also block the polymerase.

Because of these types of damage, the majority of surviving DNA is generally short—less than 100 base pairs (bp) in length
( 19 )
—and contains damaged bases. The extent of this damage is highly sample-dependent and linked to preservation conditions
( 20 )
. Cold, dry, temperature-stable environments such as permafrost regions and caves are among the best sources of well-preserved specimens and have permitted large-scale population studies
( 10,

11, 21
) . From these environments, reasonably well-preserved specimens with low levels of contamination have yielded exceptional amounts of data when next generation sequencing techniques are applied
( 22
) . As techniques improve, large-scale studies are progressively more attainable.

1.1.2. DNA Survival

Repor
ts of antediluvian DNA ( 23
) , that is, sequences greater and Antediluvian DNA

than one million years old, remain unsubstantiated and highly criticized. Particularly in the early days of aDNA research, such reports garnered much attention and publication in high-ranking journals. However, none of these claims have been independently substantiated through replication and many have been shown to be artifactual (for r
eview ( 24– 26
) ). The theoretical limit to DNA preservation remains between 100,000 and

1,000,000 years, although this varies considerably between preservation environments.

1.1.3. Contamination

The most serious complication of aDNA research stems from the small proportion of surviving copies of endogenous DNA in an extract, compared to the ubiquitous nature of environmental DNA.

The high sensitivity of PCR allows amplifi cation to proceed from only one or a few starting copies of the target sequence, but also often allows contaminating DNA to be amplifi ed. Even if the level of contamination is extremely low, PCR will preferentially amplify modern DNA over damaged ancient molecules. For example, copies of the targeted fragment may contain blocking lesions or simply be in low abundance, so that it enters the exponential phase of the PCR many cycles after the reaction has begun. If only a few contaminant molecules are present and amplifi ed during the initial cycles of the PCR, these will rapidly outnumber (and outcompete) any amplifi cation of the aDNA.

Contamination may occur at many stages of processing an

aDNA sample. The sample itself may be contaminated. For example, bones and teeth are porous, and contamination may occur via adherence or uptake of exogenous DNA, often from microorganisms residing in the depositional environment. Contamination may also occur during collection; this is a particular problem for human 4

T.L. Fulton

and microbial studies. Contamination may also be introduced during either the DNA extraction or amplifi cation processes. Laboratory personnel may introduce their own DNA or any DNA carried into the lab, reagents may be contaminated with human or animal DNA
( 27 )
, and airborne particulates may enter through the building air supply. Previously amplifi ed DNA present in the laboratory environment is a particularly dangerous source of contaminating DNA.

Even the tiny amount of DNA that is aerosolized when a tube is opened is likely to contain over a million copies of template in a volume as small as 0.005 μ L. This is potentially thousands of times more copies than all the DNA that remains in an ancient sample
( 26 )
. Therefore, it is crucial to maintain strict separation between the laboratory in which ancient samples are prepared and the post-PCR area.

Of course, not all samples share the same potential for contamination. Studies of ancient humans or microorganisms are at highest risk for contamination due to the pervasive nature of both potential contaminants. Ancient human sequences are not likely to differ substantially from modern humans, making identifi cation of contaminants nearly impossible. Bacteria, and in particular environmental bacteria, are as yet so poorly characterized that any novel sequence isolated may simply represent an uncharacterized lineage. Thus, risk assessment at the outset of any aDNA project is critical, and study design must consider contamination potential as well as the information potential from the targeted data
( 28 )
.

2. Guidelines for

aDNA Research

Before discussing the guidelines that have been proposed for aDNA research, it is important to note that following these guidelines as a mere checklist will never guarantee that the sequences produced are authentic to the samples from which they were extracted. The burden is placed upon the researcher to critically analyze the project design, to assess which criteria are perti-nent and, more importantly, to determine whether the results obtained from the experiment make sense, both in an evolutionar
y and experimental context ( 25, 28
) . Not every study must comply with all of the criteria presented below to be credible. For example, results from associated remains may not always be available or an assessment of biochemical preservation may not be necessary if the sample appears to be well preserved and the results are replicable and sensible. Above all, scientifi c judgment and rigor should prevail.

1 Setting Up an Ancient DNA Laboratory

5

Building upon pr
evious guidelines ( 3, 29
) , nine criteria for authenticity were set out roughly a decade ago by Cooper and Poinar
( 30 )
:

1.
Physical isolation of the pre-PCR aDNA facility and strict mainte-nance of a “one-way” rule of movement up the concentration gradient
: All reagents and equipment must only move in the direction of the pre-PCR facility (aka the “clean” laboratory) to the post-PCR facility. In many labs, additional precautions are taken so that once laboratory personnel have entered any building in which PCR is performed, they can only reenter the pre-PCR

facility after fully showering and changing clothes (see Note 1).

2.
Negative extraction and PCR controls
: Extraction and PCR

controls containing no DNA (negative controls) must be carried out alongside the sample(s) at every step. Positive controls (those containing DNA that have been shown previously to be successful) should be avoided due to the risk of cross-contamination. If using a positive control cannot be avoided (e.g., if a problem with a component of the extraction or amplifi cation is suspected), previously successful
ancient
specimens should be used in place of any modern specimens.

3.
Appropriate molecular behavior
: An inverse relationship should be observed between the length of the targeted PCR fragment and the strength of the amplifi cation. If very long fragments are amplifi ed as readily as are shorter fragments, it is likely that the product is a modern contaminant.

4.
Reproducibility
: Within-lab replication of PCR amplifi cations, overlapping PCR products, and amplifi cations from multiple DNA extractions must be consistent. If differences occur, more replicates should be performed.

5.
Cloning
: At minimum, a subset of PCR amplifi cations (i.e., 10%, including all unusual results) should generally be cloned to assess damage, detect nuclear mitochondrial insertions (often called numts), chimeric sequences from jumping PCR, and

contaminants.

6.
Independent replication
: A second, independent laboratory should be able to replicate any results that are obtained. Ideally, the specimen would be divided upon collection and sent to two separate facilities to avoid transfer of one lab’s potential contaminants to the other. However, if the original specimen is contaminated, this will be faithfully replicated in both facilities.

7.
Biochemical preservation
: An assessment of the likelihood of DNA preservation may be performed, for example a test of

amino acid racemization (
( 31 )
,
but see
( 32
) ), mass spectrometry to determine the peptide to single amino acid ratio, bone histology, damage determination via gas chromatography or

mass spectrometry, and bone porosity/density
( 25 )
.

6

T.L. Fulton

8.
Quantitation of starting material
: If very few surviving DNA molecules are present, stochasticity in the amount or type of damage present in the starting molecules of PCR, and PCR

error in early cycles, may produce sequence errors that appear in the majority of clones
( 33 )
. Thus, more than two amplifi cations and preferably multiple extractions should be performed.

9.
DNA from associated remains
: Particularly for high-risk studies such as those on humans, DNA preservation from animal or

other remains associated with the fi nd lends confi dence that the site conditions are conducive to DNA preservation.

These criteria have been subsequently refi ned and expanded.

Additional criteria include the following:

10.
Use of a “carrier DNA” negative
: Additional controls containing nonamplifi able “carrier DNA” should also be included.

Sometimes contaminants may be in such low concentration

that they bind to plasticware and do not amplify. However, the presence of any other DNA (such as the target DNA) may

carry them through the reaction, resulting in amplifi cation of the contaminant molecules in the PCR tubes containing the

sample DNA but misleadingly clean negative contr
ols ( 29 )
.

Including carrier DNA, such as nontarget DNA from a different source (see Note 2), with the sample may also help to allow amplifi cation of very low copy target DNA.

11.
Time-dependent or preservation-dependent pattern of DNA damage and sequence diversity
( 26, 34
) : Sequences isolated from badly preserved samples should be more damaged than

better preserved samples (as assessed via cloning or high-throughput sequencing).

12.
Phylogenetic sense
( 29 )
or otherwise reasonable results
( 28
) : Critical assessment of the sensibility of the results obtained from an aDNA experiment is an important aspect of aDNA research.

Although the sequence may be expected to be different from any known sequence, the results should be reasonable based on the genomic regions targeted and the questions asked. If the sequence is highly divergent from any known sequence, that result should make sense given the taxon being studied and the data available for comparison. For example, BLAST searching should be used to ensure that the sequences are neither human nor environmental contaminants when another species is expected.

3. Setting Up

an Ancient DNA

Laboratory

Appropriate setup of the aDNA workspace is critical if contamina-

3.1. Setting Up

tion is to be avoided. The aDNA facility should be isolated from
the Ancient DNA

any location where PCR is routinely performed, preferably in a
Workspace

separate building that does not house any PCR labs. Ideally, the 1 Setting Up an Ancient DNA Laboratory

7

room will be positively pressurized, so that air does not fl ow in from the adjoining room/hallway when the door is opened. A laminar fl ow hood or glove box provides a clean space for PCR setup, even if work is performed in dead space (no air fl ow). Reagents and equipment should never be taken into the aDNA workspace from a post-PCR facility.

When planning the layout of an aDNA lab, it is important to consider what experimental protocols will be performed in that facility. This will help to identify the amount of space required and determine whether space needs to be allocated to large pieces of equipment, such as freezers or large centrifuges. How many people are anticipated to work in the facility at a time? Can the sample preparation and extraction area be physically separated from the PCR preparation area? Have future increases in storage needs for frozen, refrigerated, or room temperature items, including samples, been considered? As in any lab, a brightly lit, highly organized facility with very little clutter on the benches is more likely to create an atmosphere conducive to the careful, precise work that is required in aDNA research.

Because everything must be newly purchased for the aDNA

workspace, it is often useful to envision a “walk-through” of the procedures that will be performed. Many items, such as paper, writing utensils, cleaning supplies, or glassware, are taken for granted in established labs. It is very inconvenient and sometimes quite diffi cult to temporarily delay a protocol, so preparation is key. Consider a solution that must be made. Is a pen and paper or calculator for recipe calculations available? Measuring devices (graduated cylinders, pipettes, etc.) of appropriate size for each ingredient, a scale that is suffi ciently sensitive to measure the required dry reagents, weigh boats, and scoops, are only a few items that should be considered. How will the solution be mixed—is a stirrer or hot plate required? If so, are stir bars of an appropriate size available in the lab? How will the solution be pH’d, if necessary? Is an appropriate storage container for the fi nal solution available? How will all of the materials be sterilized before and after use? The time it takes to ask these questions is more than made up for when a procedure can be performed smoothly without requiring a break for several days as a forgotten reagent or piece of equipment is shipped.