Analysis of trichothecene mycotoxins by mass spectrometry and tandem mass spectrometry

oo40-4020/89

Tetrahedron Vol. 45, No. 8, PP. 2251 lo 2262, 1989 Printed in Great Britain.

Analysis

of Trichothecene

Mycotoxins

by Mass

Spectrometry

and Tandem

Mass

Spectrometry Ronald

D. Plattner,

Marian

Northern

N. Beremand

Regional

and Richard

Research

G. Powell

Center

USDA-ARS 1815 N. University Peoria,

St.

IL 61604

(Received in South Africa 16 June 1988) abstract To illustrate tandem mass spectrometry (MS/MS) as an alternative to GC/MS for the analysis of trichothecenes in (DAS) levels were solid and liquid cultures, diacetoxyscerpenol GC/MS had the measured by GC/MS and by three MS/MS protocols. lowest detection limit and better precision than the MS/MS procedures; however, GC/MS required 20-30 minutes per MS/MS was more rapid, requiring less than two analysis. DAS could be measured accurately at low minutes per analysis. nanogram levels directly from filtrates of liquid cultures. Extraction of culture filtrates simplified the identification by effectively removing an interfering matrix component which Use had an ion at the same m/z as the protonated DAS molecule. of the GC as an inlet for MS/MS was an effective way to minimize physical manipulations necessary for sample All three MS/MS procedures had similar detection introduction. Occurrence levels of three limits and precisions. and DAS were measured trichothecenes, T-2 toxin, neosolaniol, in a single analysis by sequential selection of parents for each trichothecene. INTRODUCTION mass

Historically, of trichothecene technical

high

instrumentation. improvement cost

spectrometry

mycotoxins.

difficulties

the relatively

of choice mycotoxins.

to identify

they

often

Analysis mycotoxins when

of trichothecenes

produce

called

"chemical

interfering

is the limiting

of the

noise".

a rapid reductions

spectrometer.

the high target

factor

Today

mycotoxins

in the

MS is often

and quantitation selectivity

An alternative

is complicated

of

and sensitivity

at the levels

in

signals

MS is generally the target

mycotoxin,

prior

to ionization

method

of reducing

with

chemical

these

in the

signals

are

not sensitivity,

mycotoxin

by MS.

gas chromatography

or a derivative,

in the mass

2251

noise,

a particular

coupled

matrix.

constituents

spectrum:

chemical

of

sample

extraneous

in the mass cases,

by the fact that these

in a complex

spectrometer,

in the analysis

to separate

sample

MS

components

In most

in practice,

(GC) in order rest

by

only trace

into the mass

sample

Therefore,

MS offers

and to

the necessary

has witnessed

for the identification

and quantitate

to trichothecenes

and significant

a mass

for analysis

to various

occur.

are often

introduced

decade

used

both

and maintaining

capabilities

and supporting

trichothecene required

the past

been

in part,

the technique

of obtaining

However,

the technique

which

(MS) has rarely

This was due,

in applying

cost

in instrument

of obtaining

s3.w+.oo

Pergamon Press plc

from the

spectrometer.

noise,

when

a mycotoxin

is

2252

R. D. PLATTNERet al. not amenable to GC, is tandem mass spectrometry (MS/MS). MS/MS has rapidly gained acceptance in the analytical community over the past few years, and instruments with MS/MS capabilities are now available in many laboratories. The MS/MS technique had its roots in structure elucidation problems and in studies of ion structure using metastable decompositions. However, the rapid rise in popularity of MS/MS has been primarily due to recognition that the technique is particularly well suited for providing sensitive and selective analyses of complex mixtures. There are several excellent reviews describing MS/MS and its potential for the determination of trace components in complex mixtures (l-5). In a typical MS/MS procedure, a crude mycotoxin-containing sample is introduced into the mass spectrometer source and ionized from a solids probe. Mass spectra of all components present in the crude sample are thus generated simultaneously. The mycotoxin of interest is separated from most other components in the sample matrix by selection of a structurally significant ion of the target compound with the first mass filter. This selected ion is then induced to decompose by collision with a neutral target gas (usually argon) in a collision cell. Daughter ions produced by these conditions may then be mass analyzed with the second mass filter, and the resulting daughter spectrum can be used to identify or to quantitate the target mycotoxin. The signal recorded for a MS/MS daughter ion is inherently of lower intensity than would be recorded for the parent ion in conventional MS, however, interference from chemical noise in the MS/MS scan is generally greatly reduced. Because the detection limit for MS analysis with the solids probe is often limited by chemical noise, overall sensitivity, defined as signal-to-noise ratio for the analysis, is frequently improved using MS/MS. This is particularly true in cases where the mycotoxin in question is only a trace component in the sample analyzed. Detection of the mycotoxins zearalenone and deoxynivalenol at sub part-per-million levels in Fusarium-infected grains using probe inlet MS/MS has been reported (6). The only sample clean-up involved in the analysis was a solvent extraction of the grain sample. Similarly, aflatoxins have been detected in simple grain extracts at the low part-per-billion level (7). Three basic elements are common to all quantitative MS analyses; these are sensitivity, selectivity and time required for completion of the analysis. Tandem MS experiments offer the analyst a great deal of freedom in selecting from among many possible ways of performing a targeted analysis and allows the analyst to optimize any specific analysis with respect to these three elements. In developing an MS/MS targeted analysis the operator makes selective trade-offs between these three elements to obtain an acceptable blend of speed, sensitivity and selectivity appropriate to the problem at hand. In this paper, selected mycotoxin analysis problems from the authors' laboratory are used to illustrate available choices and how the variables stated above can affect each analysis.

2253

Analysisoftrichothecenemycotoxins

RESULTS

AND DISCUSSION

Studies several levels

of trichothecene

produced

approach

uses

sambucinum) well

W-induced

mutant

biosynthesize

numbers

fungal

liquid

cultures,

altered

of Fusarium

in its ability

conventional First,

culture

a rapid

were

mycelia.

medium

analyses grain

analysis. ethers

will

examined

MB1716.

instead,

in a

MB1716

is

accumulates

were

determined

MS protocols MS/MS

were

acetate,

by

examined.

analysis

work-up.

to one microliter MS/MS.

were

of one

Secondly,

the

and aliquots

of the

of the initial

Thirdly,

samples

cultures

of Fusarium

occurring

Conversion

culture

were measured

produced

simple

simplify

contaminated

mixture

by

almost

Thus,

cultures,

for each

specific

in the

from the fungal

requirements it is likely

which

cultures.

and secondary

and naturally

requirements,

grain

exclusively

analytical

it

higher

Furthermore,

in liquid

of primary

by filtration

cultures.

species

at much

the fungus.

array

are found

solid

will

for that

contaminated lead the

analysis.

TRICHOTHECENES

trichothecenes

GC results,

samples

of free hydroxyl

results

with

concentrated

different

choices

often

in naturally

infected

a complex

greatly

cultures, have

for best

generally

ethyl

and liquid

of liquid

OF DERIVATIZED

naturally

and,

with

encountered

produce

factors

to different

Most

were

3299 that has been

levels

sample

may be a relatively

analysis

of liquid

GC AND GC/MS

groups

solid

and are easily

These

samples

analyst

with

but trichothecenes

trichothecene

and,

by a direct

that trichothecenes

may still

media

tandem

without

are only partially

metabolites, culture

measured

extracted

are usually

the culture The fungus

in

methods

procedure.

In our studies

that

T-2 toxin

Three

equivalent

than

protocols

Trichothecene

by direct

levels

a

we levels

analytical

NRRL

filtrates

were

GC/MS/MS

samples

in which

snorotrichioides

extracts

be noted

created

in large

trichothecene

appropriate

as

of

to have

present

by strain

analyzed

should

levels

was conducted

analytical

(9).

were

of culture

acetate

ability

(DAS) production

GC and by GC/MS.

filtrates

filtrate

production,

the generation

approaches

for measuring

to select

to produce

(DAS)

trichothecenes

microliter

(Fusarium

has been in their

One

species.

studies.

diacetoxyscerpenol

ethyl

in order

MS- and MS/MS-based

strain

study

trichothecene

of trichothecene

trichothecene

methods

of diacetoxyscerpenol

a mutant

Pusarium

Both of these

A preliminary

encompass

to measure

uicaris

approach

are altered

(9). about

MS-based

for use in these

Several

that

information

several

of several

for Gibberella

A second

strains

of samples.

evaluated

in our laboratory methods

the inheritance

(8).

trichothecenes

need to obtain

rapid

cultures

system

to investigate traits

biosynthesis

require

in liquid a genetic

as other

study

which

approaches,

in compounds

have

should

groups

one or more

free hydroxyl

be derivatized

to trimethylsilyl

of suitable

stability

and

prior (TMS)

to

2254

R. D. PLATTNER et al.

chromatographic

derivatives a modern

with

are easily

Since

other

of the target

with

positive

the outlet

produces

the one-dimensional

Thus

GC/MS

superior

for MS, produces la shows

fairly

the EI mass

ion is observed

useful

higher

290; these

complicated

spectrum

mass

ions,

to a mass a mass

spectra

spectrum,

to deal with

are m/z combined,

378

GC detectors. ionization

of DAS.

(M-60),

The three

m/z 350

only about

Figure

spectra

1.

Electron

ionization

of the TMS derivative

(a) and chemical of DAS.

(M-60,-28),

5% of the total

ionization

are

most

intensity.

Figure

from mode

NO molecular

of ions produced

fragments.

carry

each

rather

interference

for trichothecenes.

proportion

lower mass

interfere

spectrometer,

of conventional

of the TM.9 derivative

useful

fragments

three

can severely

(EI), the traditional

The greater

at m/z 438.

the less diagnostically

m/z

response

impact

of the

is a universal

from the GC near the elution

response,

to GC in its ability

Electron

background.

onto

column.

trichothecenes.

of the GC column

intensity

injected

capillary

by the complexity

components

of the various

when

silica

detector

may elute

a multidimensional

than

is

are affected

and these

levels

a fused

ionization

components

compound

identification

By coupling component

however,

with

may be analyzed

Trichothecene

detector.

at low nanogram

equipped

the flame

matrix

time

of trichothecenes

ionization

detected

of detection,

matrix.

detector,

sample

a flame

gas chromatograph

The limits sample

TMS ethers

properties.

by GC and detected

(b) mass

and ion

Analysisoftrichothecenemycotoxins

2255

For identification of unknown components, complete mass spectra are usually recorded. However, several approaches are possible for quantitative analyses. Full scans may be acquired and the total mass spectrum at the appropriate GC retention time can be compared with a standard spectrum for positive identifications. This approach provides the greatest certainty for positive identifications. Because the mass spectrometer is continually scanned across the entire mass range for this method, only a small portion of the total analysis time is actually spent measuring the most important mass-to-charge ratios (m/z 290, 350, and 378 for DAS); therefore, this method is relativly low in sensitivity. Many mass spectrometers can also operate in the selected ion mode (SIM). In the SIM mode, the mass spectrometer only measures signals at selected m/z values. The SIM technique improves sensitivity of ion abundance measurements (typically by a factor of 50 to 100) but the procedure also has its negative aspects. In SIM, instead of a complete mass spectrum for identity confirmation of the mycotoxin, only relative intensities of the selected ibns are available. It is debatable how many correct ion ratios are necessary to positively identify a component by GC/MS, but three ions are generally regarded as a sufficient number (10). Usable full-scan EI spectra of TMS derivatives of deoxynivalenol, DAS, neosolaniol and T-2 toxin require l-20 ng of the compound. In the SIM mode, DAS can be positively identified and quantitated in the 50-100 pg range. Liquid cultures of Fusarium sworotrichioides mutant MB1716 accumulate DAS in the medium. GC/MS analysis of an aliguot of derivatized ethyl

acetate

extract

equivalent

to 1 microliter

of culture

filtrate

gave

down to a few ng/ul. DAS was quantitated by using the response of three ions, m/z 378, 350, and 290 and a calibration curve prepared from injections of known amounts of pure DAS-TMS derivative. The use of SIM to improve sensitivity was not required. Typical coefficients of variation obtained with this procedure were on the order of 7-15% and DAS levels down to a few ng/pl could be routinely quantitated. Interference from the matrix increased considerably with extracts of solid cultures and was dependent on substrate. This increase in chemical noise resulted in poorer full-scan performance. Injection of an aliquot equivalent to 10 mg of solid culture using SIM (m/z 378, 350, and 290) provided routine quantitation of DAS down to the 10 rig/glevel. reasonable

full-scan

spectra

for DAS at levels

Chemical ionization (CI) is an alternate form of ionization available on many mass spectrometers. CI is a softer ionization mode that is accomplished by collision of sample molecules with a reagent gas ion; this procedure protonates many organic molecules. For many compounds, little or no fragmentation is observed. However, trichothecenes do produce varying amounts of fragmention in CI depending on their structures. Most fragments arise from loss of substituents as neutral molecules; but a substantial portion of the total ion current is normally carried by the protonated

molecule

the TMS derivative

(MH+). of DAS.

Figure

lb shows

The MH' at m/z

an isobutane

CI spectrum

of

439 and the fragment ion at

R. D. PLATTNER el al.

2256 m/z

379 account

ionization

mode,

for most

of the

chemical

noise

fragments

and,

responses

for CI and EI are approximately

portion mass

therefore,

of the total

fragments,

Overall

with

CI.

limits

equivalent With

solid

to or better culture

that

SIM for measurement

OF CULTURE

Tandem conduct

sample

MS has the potential

but

analysis

Figure

2.

mode

since

ion a greater

diagnostic

are somewhat EI, can easily

CI identification

can be accomplished

samples

analyses

times

Chemical spectrum

to greatly

High

higher better. be and

at levels

may rapidly

mass

of of

be inserted

of a minute

spectrum

and total

or so.

of DAS

(a) and its MS/MS (b).

387

217 749

325

per

into the mass

probe,

molecule

‘I/./

to

separation

the solids

367, the protonated

16w.8-

required

for analysis

from 20-30 minutes

are made using

ionization

times

method resolution

GC requires

are on the order

of m/z

reduce

a sensitive

is time.

by capillary

when

sample

but,

in the EI mode.

provides

its cost

In contrast,

spectrometer,

daughter

GC/MS

components

analysis.

than

samples,

fewer

Total

FILTRATES

analyses.

trichothecenes,

equivalent

in CI are the more

of DAS in the full-scan

MS/MS

problem.

for trichotheoenes

quantitation require

CI is a softer

also produces

a less severe

ions generated

performance,

achieved

is often

detection

Since

ion current. from the matrix

i

-

54144.

-

218%

2257

Analysisoftrichothecene mycotoxins

Trichothecenes MS/MS

produce

of MB1716 Figure

by direct

2a shows

daughter

acid

This

ion.

fragmentation

of water

spectrometer

and then

The

ionized.

collided

(1 mtorr

with

argon

analyzed

produces Figure

by 43.

a profile 3b shows

culture

filtrate

the selected same time

Figure

daughter

MS/MS

3.

of m/z

as a pure

profile

from

188.8 1

A full daughter

scan

with

while

full daughter

different.

filtrate

that

fractional measuring standards.

8) total

intensity

are

each second, is heated. of

The profile

3a elutes

at about

of

the

of DAS.

of m/z

367 from

of strain

daughter

at this

spectrum

1 microliter

MB1716.

A)

ion response.

spectrum

than

eluting

(figure

daughter

material

probe.

307 daughter

(figure

(figure

fragment

in this

component

367 in isobutane

DAS and was

4) is virtually

standard

of an unknown

from the solids of the m/z

time

from the pure

scan of later

less volatile

only

2304.

also has an ion at m/z

distillation the

daughters

of MB1716.

of daughters

obtained

a daughter

is the daughter

is fortuitously

is set to pass

from a 1 ul aliquot

in figure

sample

is placed

ions are then

of 367 as the probe obtained

M/z. 307 is not an intense

The latter

filter

1

identical the

filtrate

and the resulting

four day old culture

307 daughter.

from

into the mass

mass-selected

daughters

(m/z 307) shown

elution

filtrate

molecules

to the

43 from m/z 90 to 400, once

ion profile

and acetic

from consecutive

of culture

from a four day old culture

(temperature)

of culture response

of eluting

the total

MS

Intense

of DAS itself.

first mass

and -2OV),

Scanning

349),

similar

inserted

These

filtrate.

acid as neutral

is quite

to dryness,

ion of DAS.

mass

and acetic

a 1 ul aliquot

m/r. 367, the parent

(at m/z

in

cultures

2b is the tandem

(m/z 367).

are observed

in the CI spectrum

evaporated

of culture

DAS molecule

pattern

can be used in liquid

of DAS and figure

fragments

daughter

MS analysis,

vial,

of an aliquot

spectrum

from the loss of water

observed

For tandem in a probe

analysis

Additional

307).

of combinations

the parent

Mli+ ions in CI that

DAS can be determined

of the protonated

are observed

(at m/z

losses

MS/MS

the CI mass

spectrum

daughters

abundant

For example,

experiments.

CI.

separated

Zb),

5) is quite spectrum.

of the culture This

component

from DAS by

DAS may be guantitated signal

and comparing

by it to

R. D. PLATTNERet al.

2258

Figure

MS/MS daughter

4.

1W.B

spectrum

of scan 54 of analysis

shown

in fig. 3.

10912.

1

1 Figure

MS/MS

5.

daughter

spectrum

se.0 187 4

OF CULTURE

Ethyl of the

acetate

later

extracted

that

FILTRATE

into ethyl

obtained

EXTRACTS

of culture

component

because

acetate.

This

Response by probe

the parent

ions generated

MS/MS

required

about

experiments

obtaining specific

(SRM), because

of the parent daughter

50-100

by measuring

full daughter

monitoring

selected

scans.

is analogous both

a parent

(MB+, m/z

ng of DAS.

and a daughter

spectra

with

daughters

were

sensitive

to detect

good

obtained

at m/z

signal-to-noise

DAS in liquid

incubation

(i.e. when

toxin

ul aliquot

of culture

filtrate

cultures

accumulation extract.

daughter

can be improved ions rather

called

selected MS:

reaction

ion are defined.

This method

With

SRM

and 229;

ratios was

after

had just begun)

in

than

it is more

349,307,289,247,

of MB1716

than

is not

only

Full-scan

and the proper

at the 10 ng level.

lower

367) represented

daughter

This procedure,

most

simplifies

was

in MS/MS

for DAS.

to SIM in conventional

(m/z 367) and daughters

step

procedure

Sensitivity

selected

removes

is not efficiently

ion transmission

in the source only

effectively

purification

this MS/MS

MS, both because

100% and because 10% of the

filtrates

this contaminant

simple

for DAS using

about

spectra

in fig. 3

I

extraction

eluting

the analysis.

shown

20s

IWE-

MS/MS

of scan 70 of analysis

of all

sufficiently

24 hours

using

only a l-2

Analysis oftrichothecene mycotoxins

2259

GC/MS/MS OF DERIVATIZED CULTURE FILTRATE EXTRACTS MS/MS analysis of culture filtrate extracts using the solids probe is simple and rapid enough to be useful for analysis of large numbers of samples but analysis of samples using the solids probe is fairly tedious. The operator must place each sample vial on the probe, mount it, evacuate the inlet, open a valve, slide the probe into the source, heat the sample, cool the probe, slide it out and close the valve for each sample (one analysis requires about two minutes). A sample inlet system that requires fewer physical manipulations would greatly simplify this procedure. One alternative is use of the GC as the sample inlet system for MS/MS. Since MS/MS is being used to detect DAS, complete separation on the column is not an important factor for identification and, therefore, short columns and higher temperatures can be used to introduce samples into the mass spectrometer source rapidly. DAS could be introduced via the GC into the mass spectrometer for MS/MS without derivatization, however, reproducability problems and losses were encountered in preliminary experiments with underivatized T-Z toxin and neosolaneol. Therefore, TMS derivatives were chosen for MS/MS experiments using the GC inlet in order to avoid problems. Figure 6 shows the CI spectrum of DAS (6a) and the daughter spectrum of the MH+ (6b). Figure 6.

MS/MS daughter spectrum of DAS.

the TMS derivative

1ea.B

1

of m/z 439, the protonated

molecule

of

R. D. PLATTNER et al.

2260 As

in

arises closely MS/l@

the daughter

from a neutral parallels

the parent, sample.

of underivatized

the CI spectrum.

experiments

Acceptable

spectrum

loss of acetic

was slightly

m/z 439, carried

acid

Response

better

than

a higher

full scan daughter

were

daughter

typical

ratios

signals

(m/2439)

for three ratio

daughter;

this

lowered

but at the expense

Figure

7.

injections

MS/MS

to measure

the ,detection

on 20-50

379,319,289,

ionization. ng of

and 229 gave the Figure

of the protonated

was greater

than

The signal

30/l.

SRM

only the m/z 439 parent limit

about

an order

7 shows

DAS molecule

of 25 ng and 250 ng.

for the 25 ng injection

may be modified

of the total

obtained

in the

DA.9 because

of a few nanograms.

379 daughter

replicate

for the TMS derivative

at m/z

injections

for the m/z

to noise conditions

from

daughter

spectrum

for underivatized

percentage

spectra

SRM of m/z 439 and daughters

proper

DAS, the major

and the daughter

to m/z

379

of magnitude,

of selectivity.

response

for six injections

for the m/z 379 daughter

of DAS

of the parent

m/z 439

(3x 25ng and 3x 250nq).

lee.

I I_

373

v

6i0

Data presented in cultures analyze

above

was

of the mutant

several

different

for the detection

MB1716.

NRRL

3299,

derived,

high

levels

and neosolaniol. trichothecenes protonated neosolaniol fragment

than

scan,

of these

detection

MB1716.

MS/MS

three

limits

provided

trichothecenes

and precisions

was

levels

for these

selecting

In chemical

of DAS

three

each

ionization

of the MS,

and as the TMS derivatives)

Therefore,

the protonated

MB1716

as lower

by alternatively

(underivatized

DAS.

as well

to

Fusarium

from which

3299 may be analyzed

for MS/MS.

DAS,

may also be used

simultaneously. strain

the yield

and thus the response

None-the-less,

analysis

analysis

of one trichothecene,

MS/MS

of T-2 toxin,

as parents

10 fold less when

scan. with

more

the parent

of NRRL

in a single

and T-2 toxin

much

for the MS/MS about

Cultures

molecules

However,

trichothecenes

SDgrotrichioides, produces

m0e

6:40

for these

molecule a useable

method

are

for the MS/MS

for simultaneous

cultures

to those

both

molecules

trichothecenes,

is selected

in liquid similar

of protonated

of NRRL

obtained

3299

for DAS in

2261

Analysisoftrichothecenemycotoxins

STRENGTHS

AND

MS/MS

LIMITATIONS

can be a more

limitations

must

Daughter

for the target

mycotoxin.

fragment

is not a common

structural

relatively

large

amount

establish

responds

the same way

are most

selected

when

of other

are quantitated

of ionization

used

is introduced

that the parent

or internal

of this

selecting

an ion

scan to insure and that

trichothecenes

by MS/MS

ideal.

for the target matrix

must be added

it

or

it must be

are far from

sample

standards

spectra

be specific

into the ion source

ion selected

in the presence

must

trichothecene

related

in

full daughter

to the daughter

daughter

other

to be valid

when

for quantitation

to the target

compounds

of matrix

standard,

be shown

be exercised

that

from components

reliable

for a MS/MS

fragment

the conditions

one must

the reference

must

is unique

When

components. that

procedure

molecule

that the parent

remembered

may arise

Care must

than the protonated

matrix

filter

several

of the possibility

they are identical

ion(s)

however,

because

mass

and when

are recorded,

to GC/MS,

First,

Identifications

analyzed.

of standards.

other

alternative

a MS/MS

compound,

CC/MS

VS.

in mind.

in the first

than the target the matrix

rapid

be kept

the ion selected

spectra

OF MS/MS

A

and

component

as it did in for reliable

quantitation.

MS/MS culture

experiments

spectrometer

and the time

Analysis

of aliquots

analysis

of culture

provide

reliable

results

agreed

(typically

analysis

with

internal

quantitative

data,

methods

using

is not necessary.

were

precision. screening

to external

direct found to

standards.

MS/MS of

was slightly

GC/MS.

Under

poorer

these

MS/MS

for accurate

of an internal

The external

was

Coefficient

standard8 with

not necessary

type method

by either

extracts,

of analysis.

external

but the addition

for a rapid

acetate

than was observed

standards

improve

in the mass

full GC separation

data compared

analyses

for DAS in liquid

to l-5 ~1 of cultures,

other

20-30%)

method

from the matrix

or of ethyl

quantitative

conditions,

suitable

consuming

filtrates

well

about

significantly

a rapid

equivalent

for MS/MS

variation

provide

DAS can be separated

samples.

standard

standard

requiring

could

procedure

only

is

l-2 minutes

per

analysis.

Mutants

were

inhibition mutant

selected

enzyme-linked

cultures

from NRF&

in which

there

T-2 toxin

biosynthesis

(9).

production

by growing

small

CIEIA.

The MS/MS

trichothecenes, tandem have

procedure

other

MS procedure

altered

3299,

immunoassay were

than T-2,

biosynthesis

to detect

for altered

are not detected for which

and using

of measuring

by CIEIA.

screen

of

T-Z toxin

plates

as it is capable

alternative

of trichothecenes

a competitive

in the level

in microtiter

flexible

that

is a potential

changes

were screened

cultures

is more

because

was available

significant

Cultures liquid

in part,

(CIEIA)

Our rapid

for mutants

which

no antibodies

are

available. Isomeric present identical

compounds,

problems

such as monoacetoxyscerpenol

for MS/MS.

molecular

weights

The three

possible

MAS

(MAS) isomers isomers

and all show the same daughter

can

all have fragments

on

2262

R. D. PLATTNER et al.

CID of the protonated intensities separated, these

prior

compounds

Design for sample measure these

when'the

compounds are made

alternative

spectra

Therefore,

in order

to ionization,

introduction,

choices

Their

ions.

occur

of a quantitative

target

flexible

molecule.

of the daughter

MS/MS

analysis

carefully

to GC/MS

in relative must be

measure

levels

of

choices

of conditions

sample.

involves

and mass

at predetermined

only

isomers

to reliably

in the same

ionization,

differ these

analysis

levels

and correctly,

for the analysis

to detect

in a specific MS/MS

and

matrix.

can be a rapid

If

and

of trichothecenes.

EXPERIMENTAL

MS and MS/MS procedures. All MS and MS/MS experiments were performed on a Finnigan TSQ-46 triple quadrupole mass spectrometer. In chemical ionization, isobutane was the reagent gas at a pressure of 0.25 torr. Source temperature was maintained at 1OOC and electron energy was 70 eV. Argon was the target gas in the Q collision cell for CID experiments. The pressure in the collision ccl 3 was approximately 1 mtorr as measured by the Bastings gauge. The collision energy was set at either -12 or -20v. Samples were introduced either via the direct insertion probe, which was heated by the ballistic ramp heater or via the gas chromatograph. For GC/MS/MS, either a 3 foot x 2mm packed column coupled to the mass spectrometer via a glass jet separator, or a six foot DB-1 fused silica capillary directly coupled to the source, was used to inlet the samples. Samples were injected, splitless, onto the column at 250 C. Cultures, media and culture conditions. Cultures were maintained on V-8 juice agar (11) slants at 4C and as conidial suspensions in 10 to 15% LiqFid cultures were inoculated at a starting density glycerol at -70C. of 0.5 x lo5 to 1 x 10 conidia per ml of media [YEPD-5G (9)] with conidia washed from strains grown on V8 agar plates for 7 to 10 days. Liquid cultures were grown in Erlenmeyer flasks in a volume of medium equal to one-half the volume of the flasks for 6 to 7 days at 28 C on a At selected intervals, aliquots rotary shaker operated at 180 to 200 rpm. of the cultures were removed and filtered. Filtrates were stored at -20C or extracted with ethyl acetate (equal volumes, 2 times, until analyzed, combined) and stored until analyzed. For GC analyses, portions of ethyl acetate extracts were cleaned up using a charcoal column procedure (8). The clean-up procedure was omitted for GC/MS analyses. For preparation of TMS-derivatives a fraction of the sample equivalent to a l-ml portion of the original sample was evaporated to dryness at 80C under nitrogen, reacted with 100 ul of trimethyl silylating reagent (Tri-Sil/TBT; Pierce Chemical Co., Rockford, 111) for 1 hour at 8OC, and brought to 1 ml by addition of 900 pl of hexane. References John Wiley & Sons: McLafferty, F. W., Ed. "Tandem Mass Spectrometry"; New York, 1983. D. D., Mass Swectrom. Rev,_, 1983, 2, 2. Yost, R. A. and Fetterolf, l-45. G. L., Chem. Ena. News, 1981 (Nov. 30). 3. Cooks, R. G. and Glish, 40-52. F. W., Scinece, 1981, 214, 280-287. 4. McLafferty, 5. Johnson, J. V. and Yost, R. A., Anal. Chem, 1985, 57, 758-768a. Plattner, R. D. and Bennett, G. A., J. Assoc. Off. Anal. Chem., 6. 1983, 66, 1470-1477. R. D., J. Assoc. 7. Plattner, R. D., Bennett, G. A., and Stubblefield, Off. Anal. Chem., 1984, 67, 734-0. 1987, 77, Desjardins, A. E., and Beremand, M. N., Phvtowatholoay, 8. 678-683. Awolied and Environ. Microbial. 1987, 53, 1855-9. Beremand, M. N., 9. Mass Spectrometry, Aeyden, London, 10. Millard, 8. J., QuantitatG 1977. 1974, p. 691-703., University of 11. Stevens, R. B., Mycology Guidebook, Washington Press, Seattle, Washington. 1.