Michelle Metchikian 26 min

Modified DNA Extraction Procedure from Fired Cartridge Cases


Over the past nine years, the Verdugo Regional Crime Lab has performed DNA testing on over 1000 cartridge cases, live rounds, and projectiles from casework. We implemented the “soak” method (similar to the method developed by San Diego Police Department) in December 2014 and the BTmix (developed by the Alcohol, Tobacco, Firearms and Explosives Lab) + FLOQSwab method in September 2022. This presentation will cover the validations, workflows, casework statistics, challenges for both methods, and why we made the decision to modify our extraction procedure.



0:00

[MUSIC]

0:15

Hi everyone, I hope you're enjoying his 2023 as much as I am.

0:19

I especially liked seeing the rapid-hit ID instrument being driven around by my

0:24

colleagues, Peter, John, and Tamar.

0:26

I'm Jamie Brackold, the Senior Manager of the Forensic Science Applications

0:30

Group,

0:31

and I'm pleased to be introducing the next presentation,

0:35

a modified DNA extraction procedure from fired cartridge casings.

0:39

Joining me today is Michelle Michican from the Verdugo Regional Crime

0:44

Laboratory in Glendale, California, where she's a criminalist.

0:47

She'll be covering the modified extraction procedure that her lab is using and

0:51

developed for fired cartridge casings.

0:54

And she'll also talk a little bit how that has impacted workflow efficiency and

0:58

first pass success rate from these difficult sample types.

1:03

After the presentation, we will have time to answer some questions, so make

1:06

sure to enter those in the question box at the bottom of your screen.

1:10

And with that, I'll pass it over for the presentation. Thank you.

1:13

[BLANK_AUDIO]

1:17

Hello, my name is Michelle Michican, and I am a criminalist at the Verdugo

1:22

Regional Crime Lab.

1:24

Located in Glendale, California, our lab is housed within the Glendale Police

1:29

Department.

1:30

Today, I will be sharing our modified DNA extraction procedure for fired

1:34

cartridge cases.

1:36

According to the FBI Uniform Crime Report, in the US in 2019, 73.7% of murders

1:44

included the use of a firearm,

1:46

36.4% of robberies, and 27.6% of aggravated assaults.

1:53

Firearms are rarely recovered from crime scenes, but cartridges in cartridge

1:58

cases are.

1:59

Oftentimes, they are the only physical evidence recovered.

2:04

We know that touch DNA may be left behind on the cartridges when an individual

2:08

loads them into the magazine.

2:10

Due to the small surface area, the amount of DNA deposited is typically quite

2:15

minimal.

2:16

In addition, during the firing process, cell loss and degradation occurs.

2:21

As a result of all these factors, there is a need in our community to maximize

2:27

DNA recovery from cartridge cases.

2:29

The SOAP method was first introduced in a paper published in 2011 by Deal Just

2:37

Set All.

2:38

From 2003 to 2009, 26.5% of criminal cases that they were yielded DNA profiles,

2:46

and 6.9% of the case work samples yielded DNA profiles.

2:50

They defined success as any reproducible DNA results with typing results at two

2:55

or more low-side.

2:57

San Diego PD adopted this method, and from 2011 to 2014, had a 26.1% success

3:05

rate per sample.

3:07

Learning about these results, in 2014, we decided to start our SOAP study.

3:14

While Glendale, California typically only sees a handful of shootings a year.

3:18

In addition to servicing Glendale, we are a fee-for-service lab and perform DNA

3:24

testing for over 25 local, regional, and federal agencies.

3:28

This is a timeline of DNA testing of cartridge cases in our lab.

3:32

In 2014, we validated and implemented the SOAP method.

3:37

In 2021, we started our modification study, and in September of 2022, we went

3:44

online with the BT Mix and Flock SOAP method.

3:46

We processed approximately 900 case work cartridges, cartridge cases, and

3:51

projectiles using the SOAP method.

3:53

And thus far, we have processed approximately 161 case work cartridges,

3:58

cartridge cases, and projectiles using the new method.

4:01

In this PowerPoint, I will discuss the SOAP study.

4:05

Our case work results using the SOAP method.

4:07

While we made the decision to modify our procedure, our modification study, and

4:13

our case work results from the BT Mix plus Flock SOAP method.

4:17

For our SOAP study, we had volunteers handle 10 cartridges for up to two days.

4:22

Volunteers then loaded and fired their personal firearms.

4:26

The cartridge cases were collected and packaged by another volunteer.

4:30

We obtained reference samples from the handlers and the collectors.

4:35

To extract the DNA from the cartridge cases, we placed them in 5 ml tubes and

4:39

soaked them in digest buffer and pro K.

4:42

The max volume we can load on the automates, which is what we use for the

4:46

extraction, is 530 microliters.

4:49

Therefore, we did our best to add enough lysis buffer to reach the lip of the

4:54

spent cases, but not more than 1,060 microliters.

4:59

After we submerged the cases, we incubated them at 56 degrees Celsius for 30

5:04

minutes.

5:04

We then swabbed each cartridge case using a Puritan cotton swab and put the sw

5:09

ab in a prep filer column tube assembly.

5:12

We added approximately half the lysate to another column tube assembly and the

5:17

remaining lysate to the assembly with the swab.

5:20

Following the automated express extraction using the prep filer express

5:25

forensic kit,

5:26

we then combined it to extracts per sample using micro column columns.

5:30

At that time, we quanted the extracts using the quantifiler human kit,

5:35

amp them using Identifiler Plus and type them using 83500.

5:40

We processed a total of 65 cartridge cases for the soak study.

5:45

The average quant was 0.006 nanograms per microliter with the nickel cases

5:52

having a noticeably higher quant average compared to the brass cases.

5:56

13.8% of the cases yielded full DNA profiles and 13.8% yielded partial DNA

6:03

profiles.

6:04

We defined partial as having typing results at 7 to 12 codis core loci.

6:09

Coinciding with the average quant values, the nickel cases yielded higher

6:14

percentages of interpretable profiles when compared to the brass cases.

6:20

Further, as expected, the larger calibers yielded more interpretable results.

6:25

We did see four mixtures in the study and for all of them, the non-loader alle

6:31

les could be attributed to the collector.

6:33

We also obtained one single source DNA profile that matched the collector.

6:38

Since the other four 9mm cases corresponding to that loader did not yield

6:43

partial or full profiles,

6:46

the lack of the handler's DNA detected from the extract was expected.

6:50

The collectors were wearing gloves but not masks.

6:54

The volunteers were sworn personnel and when we went over the results with them

6:59

it really changed our perspective of how sensitive DNA testing is.

7:02

We also observed artifacts from multiple handlers, calibers, cartridge case

7:08

types and extraction sets.

7:10

In the almost eight years that we used a SOAP method in case work, we processed

7:14

approximately 918 cartridge cases, cartridges and projectiles.

7:19

For the first two and a half years, we used quantifiler human and identifiler

7:24

plus.

7:24

In June of 2017, we went online with quantifiler trio and global filer.

7:30

From those 918 case work samples we processed using the SOAP method,

7:35

approximately 16% of the samples were interpretable.

7:40

Approximately 47% of the 198 criminal cases we worked with these sample types

7:46

resulted in at least one interpretable profile.

7:49

One of the factors that limited the number of samples that we can process with

7:53

the SOAP method was the need to split samples during extraction

7:57

and then having to combine and concentrate them using microcons.

8:01

The automates can extract 13 samples at a time and in our lab we extract two

8:06

region blanks per extraction set.

8:09

Since we treat three region blanks in the same manner, we were only able to

8:13

extract four handgun samples per extraction set.

8:16

This became even more limiting for us when rifle calibers were submitted for

8:22

testing.

8:22

When we soaked the samples, we did our best to submerge them as much as we

8:27

reasonably could.

8:28

For handgun calibers, we would add up to approximately 1000 microliters of l

8:33

ysis buffer, but for the rifle calibers, we would add

8:37

2000 microliters and even then depending on the caliber, as you can see in this

8:42

photo, the cases were not always fully submerged.

8:46

Nonetheless, when we would add 2000 microliters of lysis buffer, we would then

8:51

need to split the lysate into four column tube assemblies to load onto the

8:55

automates.

8:56

Since we had to split each region blank into four, we only had room for one

9:01

sample per extraction set.

9:03

Another challenge that we encountered in case work, just as we saw in the

9:07

validation study, was numerous samples when noisy baselines and many artifacts.

9:12

We were not able to correlate these artifacts with any particular caliber or

9:16

cartridge case type.

9:18

We made it standard practice to replay and retype the samples anytime artifacts

9:23

were observed to determine their reproducibility and eliminate the possibility

9:28

of well contamination on the plate,

9:30

which added an additional step to the standard workflow.

9:34

Depending on the profiles, these artifacts yielded interpretation challenges.

9:38

In 2021, we decided to look into modifying the procedure as we were routinely

9:43

processing the sample types and casework and were encountering limitations and

9:48

challenges.

9:49

The main factor that we wanted to try to address was efficiency, of course,

9:54

without compromising success.

9:57

The extraction process required two incubations and the labor intensive

10:01

processes of needing to split the lysates, having to microcon to concentrate,

10:06

and routinely needing to replay samples due to the presence of artifacts.

10:11

This limited the number of samples an analyst can process in a batch.

10:15

In addition, we wanted a procedure that we could more easily automate in the

10:20

future.

10:22

We also wanted to see if there was a reagent that can be used to help mitigate

10:26

the presence of artifacts and elevated baselines.

10:29

Another factor that was considered was cost.

10:33

Could we minimize costs which went hand in hand with improving efficiency?

10:38

If we could eliminate the need to split lysates, that would result in using

10:42

pure prep filer cartridges, both for samples and reagent blanks,

10:47

and it would help eliminate the need for microcons. And if we could cut out

10:52

some of the labor,

10:53

that would free up time for analysts to work on other tasks and we all know

10:56

time is money.

10:57

Lastly, can we increase success? As techniques become more advanced, can we

11:04

adopt new ways to improve our recovery?

11:06

Further, if we can decrease manual manipulation, we would subsequently be able

11:12

to decrease contamination risk, which would aid an overall increased success.

11:17

Before I get into the procedures that we tested, I want to discuss what Flock

11:22

Swabs and BT Mix are.

11:23

Flock Swabs are sterile swabs manufactured by Copan.

11:27

According to the manufacturer, they consist of a tip that is spray coated with

11:32

nylon fibers that are arranged perpendicularly on a plastic shaft.

11:36

This allows for easier scrubbing of the surface to recover more DNA as well as

11:42

ease of release of the DNA during the digestion process.

11:46

The plastic shaft breaks off, so when we place them into the prep filer columns

11:52

, we can easily break off the tip by bending the shaft back.

11:54

For the BT Mix, we were provided the recipe by ATF who had started using it for

11:59

their casework.

12:01

In 2020, Bill et al. from ATF published a paper discussing their wash method in

12:06

the presence of BT Mix.

12:08

The B and BT Mix stands for Bovine Serum Albumin or BSA, which is known to keli

12:14

contaminants.

12:16

The T stands for Tripeptite Glycin Glycin Histidine or GGH.

12:20

GGH is the copper binding portion of BSA, so it's used to keli copper.

12:26

The presence of copper in the cartridge cases can potentially cause inhibition,

12:30

degradation, and more artifacts.

12:32

For our modification study, we had volunteers carry 16 9 millimeter cartridges

12:38

in their pockets for two days.

12:40

Volunteers then loaded and fired their personal firearms.

12:45

The cartridge cases were collected and packaged by another mass and glove

12:48

volunteer.

12:49

We obtained reference samples from the handlers and the collectors.

12:53

We evaluated three new procedures. For procedure one, we added 30 microliters

12:59

of BT Mix to a fogh swab and swab 72 cartridge cases.

13:03

We did not soak the cases.

13:06

We then extracted using the prep filer kit on the automate and eluded the

13:09

extracts in 20 microliters.

13:13

For procedure two, we used our validated soak method, but added 30 microliters

13:17

of BT Mix to the digest buffer and used the floc swab rather than a Puritan

13:22

cotton swab to swab the cases.

13:24

Additionally, rather than using microcons, we used a vacuum fuse to concentrate

13:30

the extracts.

13:31

As you can see from the number of cartridge cases tested on the table, half the

13:36

amount was processed compared to the other procedures.

13:39

We discontinued this method due to very poor results.

13:43

The floc swabs don't absorb a lot of liquids, so swabbing with them after the

13:47

cases had soaked was problematic.

13:49

For procedure three, we added 30 microliters of prep filer lysis buffer to a fl

13:54

oc swab and swab 72 cartridge cases.

13:57

We did not soak the cases.

14:00

We then extracted using the prep filer kit on the automate and eluded the

14:03

extracts in 20 microliters.

14:05

This was the same procedure as procedure one with the exception of adding prep

14:13

And lastly, we extracted 72 cartridge cases using our validated soak method so

14:17

that we can do it side by side.

14:20

To eliminate cheddar status as a factor, each analyst extracted four cartridge

14:24

cases from the same six sonars for each procedure.

14:28

We quanted the samples using quantifiler trio, amped using global filer and

14:33

typed them on a 3500.

14:36

These are the quant results from procedures one, three, and four from our

14:40

modifications study.

14:41

For procedure one, which was the method where we solved the cases with BT Mix

14:46

added to floc swabs,

14:47

the average quant was 0.033 nanograms per microliter.

14:51

For procedure three, which was the method where we solved the cases with prep

14:55

filer lysis buffer added to floc swabs,

14:58

the average quant was 0.00298 nanograms per microliter.

15:03

And for our validated soak method, the average quant was 0.00346 nanograms per

15:09

microliter.

15:10

In other words, procedure one yielded the highest average quant followed by our

15:14

validated soak method and then procedure three.

15:17

Success was determined based on the overall percentage of interpretable

15:22

profiles, regardless of whether the shooter was present.

15:24

As expected, procedure one with the highest average quant value correlated with

15:29

the highest overall percentage of interpretable profiles with 48.6% being

15:36

interpretable.

15:37

The cut-off for interpretable profiles for the modification study was

15:40

considered to be 7 loci, including a miele genon and the y markers, if

15:45

applicable,

15:46

as this is typically enough loci for comparisons.

15:49

Profiles that were considered generally consistent with the shooter with those

15:53

profiles with the majority of the alleles were consistent with the shooter and/

15:58

or the shooter is present in the mixture.

16:00

If the shooter was present in the mixture, he or she may not be the major

16:04

contributor.

16:06

To be consistent with the shooter, that meant that the overwhelming number of

16:10

alleles were consistent with the shooter's alleles.

16:13

At times, we saw the collector present in the samples.

16:16

This, like our soak study, once again demonstrates that extreme caution should

16:21

be made when handling these items in the field.

16:24

We observed one contamination event from an analyst. We also had one profile

16:30

that had no alleles attributable to the shooter.

16:32

The profile was consistent with another officer, however, this officer was not

16:37

the collector.

16:38

The officer was involved with the study, though.

16:41

It is unclear exactly how the transfer occurred or whether that officer could

16:45

have directly handled the cartridge.

16:48

For procedure 4, the soak method, one profile of those potentially interpret

16:53

able profiles may be too genetically complex due to the presence of artifacts, a

16:58

noisy baseline, and potentially two contributors.

17:02

Several alleles were consistent with the shooter, however.

17:06

For procedure 3, the average quant value was 0.00298 nanograms per microliter,

17:12

which was slightly less than procedure 4.

17:15

However, allele recovery in procedure 4 was much lower than procedure 3.

17:20

The percentage of interpretable profiles for procedure 3 was 14%, and for

17:26

procedure 4 was 5.6%.

17:29

These three profiles are from the same loader, and the cases were extracted

17:33

using the soak method.

17:34

As you can see, the first one has numerous artifacts, and some artifacts are

17:38

seen in the other two profiles.

17:40

When cases from the same loader were extracted using procedure 1, the BT-MIX-F

17:45

OC-SWA method,

17:46

you can see that there is a low baseline and no artifacts were observed.

17:51

Based on the quant data, the percentage of interpretable profiles, and being a

17:55

more cost-effective and efficient sampling and extraction procedure,

18:00

we decided to implement procedure 1, the BT-MIX-FOC-SWA method, as it met all

18:05

of our goals.

18:07

This method only requires one incubation, as opposed to two.

18:12

We no longer need to split the lysate and then combine and concentrate

18:16

following extraction.

18:18

This procedure also can be more easily automated, which is a future goal.

18:23

We are no longer seeing artifacts and elevated baselines, which has eliminated

18:27

the need to replay our samples.

18:31

Further, the lack of artifacts has eased interpretation. All these factors have

18:36

tremendously contributed to this method being more efficient,

18:39

and as a result, analysts are capable of processing more cartridge cases per

18:44

batch.

18:44

This more efficient method has cut down the overall costs per sample.

18:49

We are using fewer prep-filer cartridges, plastics, and consumables.

18:54

We no longer need pricing microphones, subsequently, since we are running fewer

18:59

reagent blanks,

19:00

we are saving on consumables and reagents post-extraction.

19:04

So while flockswabs are more expensive than pure tin swabs, and the BT-MIX

19:09

comes out to about $3 per sample,

19:12

we are still saving money with everything else we are cutting.

19:16

Not to mention, this new workflow has freed a time for analysts to work on

19:20

other tasks,

19:21

as opposed to spending longer and multiple extraction days for the same number

19:26

of samples.

19:28

And of course, we picked this method because it yielded a higher recovery rate

19:32

and percentage

19:33

of interpretable profiles. Also, since there is far less manual manipulation,

19:39

we felt there was a decreased contamination risk, which would aid an increased

19:43

success.

19:44

Currently, based on 161 case work samples that we have extracted using the BT-

19:50

MIX flockswab method,

19:52

approximately 20% have yielded interpretable profiles. Approximately 45% of the

19:58

38 criminal

19:58

cases we've processed thus far resulted in at least one interpretable profile.

20:04

Our sample size thus far is not very large, but we are definitely seeing

20:08

comparable results

20:09

to our SOAP method. I would like to thank the Forensic Biology Staff at VRCO

20:15

for all the hard work

20:16

that they put into the studies. Breenora and myself did all the lab work for

20:20

the modifications

20:21

study. Our tech lead Heather Sargent did the data analysis and completed the

20:26

write-up,

20:27

and lab director Katherine Nguyen reviewed the data and the write-up. Our lab

20:31

would like to thank

20:32

Todd Bill from ATF for providing the BT-MIX recipe to our lab. We would also

20:37

like to thank

20:38

Scott Schroeder, Dave Jackson, Kelsey Kemp, and Shurine Darwish from Thermo

20:42

Fisher Scientific

20:44

for their input and support with the studies, and I would also like to thank

20:48

Thermo Fisher

20:48

for inviting me to speak today. And we would like to thank Glendale PD,

20:53

Pasadena PD,

20:54

Anaheim PD, and San Jose PD officers for helping with the shoots.

20:58

Thank you. I'll be sticking around. If you have any questions, feel free to

21:04

drop them in the chat box.

21:05

Great presentation, Michelle. We do have some questions that came in from chat,

21:06

filer lysis buffer as opposed to BT Mix.

21:15

so let's get started on those. So the first one, have you tested swabbing fired

21:20

cartridge

21:21

casings using only DI water instead of BT-MIX? No, we have not.

21:27

All right, we have another one. Do you know if the BT-MIX affects the

21:33

composition of the cartridge

21:35

cases, such as it could affect future analysis or comparison by a firearms lab?

21:42

About half of the cartridge cases that we process using a new method go on to n

21:48

iven entry,

21:49

and we work very closely with that unit, and they typically get their entries

21:53

in within a

21:54

few days. They haven't told us anything about the quality being affected by it.

22:00

Now,

22:00

long term, that could be something to look into, but long term, I don't know

22:03

the answer.

22:04

Great to hear about the partnership between different parts of the crime lab.

22:10

Do you sometimes swab multiple cartridges with a single set of swabs, or do you

22:15

only swab

22:15

single swabs? You use single swabs, I guess. We only use single swabs, we only

22:21

swab one item

22:23

with one swab. Okay, another question, did you test method one using a cotton

22:31

swab with the BT-MIX

22:33

rather than the flax swab? No, we did not. And another question about flax sw

22:40

abs?

22:41

Flax swabs are not very absorbent. Did you encounter challenges when using them

22:46

This is a question that we have bought in a couple of times from different labs

22:51

. Yes,

22:51

flax swabs are not very absorbent. So when I'm in lab, one thing I do when I'm

22:55

dispensing the

22:57

30 microliters of BT-MIX onto them is I start moving the pipette tip around so

23:02

that it's being

23:03

dispensed kind of along the area of the swab itself. And then when I am swab

23:10

bing the items,

23:11

because they aren't absorbent, I see that there's quite a bit of liquid that is

23:17

deposited onto

23:19

the cartridge cases. And so after I do a very thorough swabbing, I then go back

23:24

and very lightly

23:25

go over the exterior of it trying to pick up all of the residual liquid.

23:31

Great. Science is definitely an art and it sounds like you have the technique

23:34

down.

23:35

Okay, another question. The SOAP method in the modification study was only 5.6%

23:44

compared to your

23:45

original study, about 26%. Anything obvious leading to the difference?

23:50

I mean, inherently, there is variability in how much we're going to recover

23:56

depending on

23:57

the shutter status of these individuals. But the one main factor that I could

24:02

think of is that in

24:03

our original study, we did process nickel cartridge cases where we saw a higher

24:09

recovery rate from

24:10

them. Whereas for our second study, we only process brass cartridge cases.

24:17

Okay, there's actually another question about brass casings I'm seeing here.

24:24

Where there's signs

24:24

of inhibition with the brass casings either at quant or amp.

24:28

At quant, there was not the IPCCTs where what we would expect for a non-in

24:35

hibited sample.

24:36

And then as far as amp, that's hard to say yes or no because the quantities are

24:42

solo to begin with.

24:43

So just inherently, there might really not be DNA there. So the amp part, we

24:49

didn't detect

24:49

anything. That's not to say there wasn't. Okay. You mentioned handling for two

24:57

days before

24:58

loading and firing the weapon. What did that entail? The officers were

25:03

instructed to carry

25:04

into 10 cartridges in their pockets. Okay, we have about a minute left. I'm not

25:13

seeing any other

25:14

questions. So I'll give everyone just a couple seconds. They have any last

25:19

minute questions to

25:20

throw them in. All right, I think we have time for just this last one. Is your

25:37

procedure online

25:39

for anyone to use or is there a way to access your lab's procedure? You can

25:44

definitely contact

25:45

our tech lead. I'm not sure exactly. You can try to get in touch with me or

25:50

reach out to

25:50

Thermo Fisher and then they can get us connected. That sounds great. All right,

25:58

well, thank you so

25:59

much, Mish. Oh, wait, we had one more come in and the nick of time. How soon

26:04

after swabbing were the

26:05

samples extracted? Like right away. That's part of the procedure. So we swab

26:12

and extract right

26:13

away. So it's not like when we're in lab and we're screening the evidence where

26:18

we'll swab it,

26:19

put it in a locker and then maybe a week later extract. No, we do it right away

26:23

. Okay. All right,

26:24

thank you, Michelle, so much for your time. It was great hearing about the new

26:27

procedure you guys

26:28

are using. Thank you guys. Okay.