You are on page 1of 7

Bacterial Identification for Environmental Isolates Utilizing MALDI-TOF

Technology

Introduction

Most technologies and databases designed for rapid identification of


microorganisms are designed with clinical isolates in mind. When adapted for
use in the microorganism monitoring programs of pharmaceutical, nutraceutical,
and sterile manufacturing environments, these technologies underperform due to
the diversity of organisms encountered in these environments. The development
of rapid and automated microbial methods and technology are largely driven by
the need to rapidly identify clinical isolates, and the strain databases that these
technologies use are invariably focused on these microorganisms.

Environmental Monitoring programs often encounter a high diversity of


organisms that originate in water and soil as well as human associated
microbiota. For industries that are required to identify microorganisms on a
regular basis, the ideal technology is one that is accurate, easy to use, is fast and
inexpensive. An additional consideration is the relevance and entries in the
library database to identify organisms to the genus/species level.

One potential technology is using matrix-assisted laser desorption/ionization


time of flight (MALDI-TOF) spectroscopy. Whole cell proteotypic analysis with
MALDI-TOF spectroscopy is an emerging rapid and inexpensive method of
identifying bacteria. MALDI TOF analyzes primarily ribosomal proteins which
are not differentially expressed making this technology less dependent on growth
conditions then phenotypic methods. This method is also faster, less
complicated to perform and lower cost than genotypic identification methods.

Identifying bacteria with MALDI-TOF (Figure 1.) a single isolated colony (A) or
simple cell extract is spotted onto a stainless steel target plate (B) and overlaid
with an ultra violet absorbing molecule. The target plate is inserted into the
MALDI-TOF (Bruker Microflex) instrument (C). Nitrogen pulsed laser ionization is
applied to the sample and the proteins are ionized. They are then separated
based on their mass/charge ratio (D). The resulting spectra, a protein finger print
(2,000 20,000 Dalton range) are compared to a database of known spectra
(Burker BioTyper) (E). Processing time is quicker than both genotypic and
biochemical technologies to obtain an identification, taking 10 minutes for the first
sample to be read with an additional 1 minute/sample thereafter. A match factor
for each sample is calculated based on the number and amplitude of the peaks
that match as well as the number of peaks that do not match the known spectra.
This study evaluates MALDI-TOF technology for its ability to correctly identify
environmental isolates.
A B

D
Ribosomal Spectra

Top: unknown spectra

Bottom: known spectra

Methods

A set of 31 known bacterial strains from culture collections and 146 different
strains representing 43 different species know to frequently occur in
Pharmaceutical and manufacturing environments were used for this study.
Identities of all organisms and strains were verified using 16S sequencing cross
referenced against the Accugenix library database.

Bacteria for each experiment were tested using both the direct spotting method
and extract method (ethanol-formic acid-TFA) per manufactures
recommendations.

The following variables were tested:


- 24 hour cultures on three different days to investigate overall variability
- 24, 48, 72 hour old cultures to determine if the age of the culture effected
identification
- Different media types were used to determine if different growth conditions
effected the ability of the organisms to be identified.
A total of 1,873 data points were obtained, the results of which are summarized
below.

Results

The overall variability of MALDI-TOF processing is higher than genotypic testing,


but there were no significant trends associated with processing samples from
different media types. Cultures greater than 48 hours old had lower match
factors overall but still identified correctly the majority of the time. Very few
organisms were misidentified. When observing the average match factors of
organisms whos closest match with the BioTyper software was correct, it was
found that the species tested could confidently be identified with a match factor
cut off of 1.75 (Fig. 2), which is lower than the software's default
recommendation, with very little risk of misidentifying organisms (Fig. 3).

Figure2: Percentage of samples that successfully identify to the correct species


(based on DNA sequencing), fail to result in an identification, and identify
incorrectly at various match factor cut offs.

Figure 3: Distribution of sample identifications with BioTyper software


recommended interpretation (A) Distribution of sample identifications with the
match factor cut off set at 1.75 for a species level identification (AccuPRO-ID
cutoff) (B).
Data Set with Manufacturers Software Recommended Interpretation
(n=1873)

20.3%
26.1%
Secure Species ID
Probable Species ID
Probable Genus ID
No ID

22.7% 30.9%

Data Set with Optimized Accugenix Interpretation (n=1873)

30.2%

Species
No ID

69.8%

Overall performance of MALDI TOF spectrometry for identifying environmental


isolates resulted in accurate species level IDs 70% of the time (Fig. 4). The error
rate was very low with only 1.8% of species being misidentified. A small
percentage of samples did not produce analyzable spectra, which is an indication
of processing variability.

Figure 4. Performance of sample identifications with the match factor cut off set
at 1.75 for a species level identification using DNA sequencing as a reference.
Overall Performance of MALDI-TOF (n=1873)
1.8%
5.9%

Species
22.5%
No ID
MisIdentification
No Spectra

69.8%

Discussion

One of the factors contributing to a large number of samples not being


successfully identified with MALDI-TOF is that some species frequently seen in
environmental monitoring programs are missing from the database of known
spectra. The inclusion of adding relevant species to the database used by
Accugenix has increased the number of isolates identified (Fig. 5).

Identification Rate with Updated Library

17.1%

Species
No ID

82.7%
The percentage of sample identifications increased from 69.8% to 82.7% with the
addition of 15 relevant species to the AccuPRO spectra database (match factor
cut off set at 1.75 for a species level identification).

Conclusions

The low cost, ease of use and rapid turnaround time of MALDI-TOF analysis are
strong advantages of this technology. A customized database optimized for
environmental monitoring of isolates commonly encountered in the
Pharmaceutical and related manufacturing industries are required for better
performance. The ability of shipping ethanol extract samples to a contract
service provider is an inexpensive proposition, as well as a technical benefit
because temperature and time dependencies are eliminated.

About Accugenix:

For over 20 years, Accugenix has been the worlds leading FDA-registered,
cGMP service laboratory for microbial identification and characterization. The
company has tested and identified more microorganisms than any other
industrial company or service laboratory, which has allowed Accugenix to
develop and refine a
validated proprietary library and method of interpretation unrivaled in
industry. More than 1000 facilities around the world rely on the companys
expertise to meet their microbial identification needs.

Try out AccuPRO-ID for free! Simply submit 5 bacterial samples and write the
redemption code: Pharmig-AD in the Comments column on the Identification
Request Form (AccuPRO-ID IRF Page 1). To access the form, please visit us at
http://accugenix.com/info_ref.php. For questions concerning sample submission
contact our Technical Support Organization at technicalsupport@accugenix.com
or call 1 302.292.8888.

Mark Calmann
Sr. Product Manager
Accugenix
223 Lake Drive
Newark, DE 19702
markcalmann@accugenix.com

You might also like