13.2 BSL-3 Pathogens and Select-Agent Presumptive Identification
Key Takeaways
- Sentinel workflow is recognize, rule out, refer: blood, sputum, tissue, and lymph-node cultures with biothreat colony or Gram-stain features are worked only in a Class II BSC.
- Bacillus anthracis: large nonhemolytic ground-glass medusa-head tenacious GPR that is catalase-positive and nonmotile. Do not place it on automated ID or MALDI.
- Yersinia pestis: slow fried-egg colonies, bipolar Gram-negative rods, oxidase-negative, urease-negative, and nonmotile at 35°C.
- Brucella spp.: tiny slow GNCB, no MacConkey growth, catalase/oxidase/urease-positive. Francisella tularensis: tiny, cysteine/CHOC better than BAP, no MAC, oxidase-negative, catalase weak or negative, beta-lactamase-positive.
- ASM August 2025: Brucella was removed as a U.S. select agent by the final rule published December 17, 2024 and effective January 16, 2025. Deregulation changes biosecurity paperwork, not biosafety; it remains highly infectious, reportable, and a rule-out/refer organism. The ASCP outline still lists it among BSL-3/select-agent examples.
13.2 BSL-3 Pathogens and Select-Agent Presumptive Identification
Quick Answer: Blood, sputum, tissue, and lymph-node specimens can yield BSL-3 agents. Sentinel laboratories recognize suspicious colony and Gram-stain features, perform a short rule-out panel inside a Class II BSC, and refer anything that cannot be excluded. Do not place these isolates on automated ID systems or open-bench MALDI-TOF. Brucella was removed from the U.S. select-agent list on December 17, 2024; biosafety and rule-out/refer duties did not change.
The ASCP M outline II.M BSL-3 Pathogens and Select Agents (Bioterrorism) still lists Bacillus anthracis, Yersinia pestis, Brucella spp., and Francisella tularensis as the examples for colony morphology and rapid tests used for presumptive identification. Your job at the bench is not to confirm a select agent. Confirmation belongs to the Laboratory Response Network (LRN) reference laboratory. Your job is to recognize the pattern, rule out look-alikes with a handful of tests that can be done safely in a BSC, and refer.
Specimen sources
These organisms arrive as ordinary-looking cultures:
- Blood (and bone marrow): Brucella, F. tularensis, B. anthracis, and Y. pestis (septicemic plague).
- Sputum / respiratory secretions: inhalational anthrax, pneumonic plague, pneumonic tularemia.
- Tissue, aspirates, eschars: cutaneous anthrax, ulceroglandular tularemia.
- Lymph node / bubo aspirate: bubonic plague, ulceroglandular tularemia, brucellosis.
Any slowly growing Gram-negative coccobacillus from blood, or a large nonhemolytic Gram-positive rod from a blood or respiratory culture, should stop the usual open-bench workflow.
Sentinel algorithm: recognize → rule out → refer
- Recognize. Match Gram stain, colony speed, hemolysis, and MacConkey growth to a biothreat pattern.
- Rule out. Move the plates into a certified Class II biological safety cabinet. Complete only the rapid tests in the ASM sentinel guideline. Use BSL-3 practices (or a BSL-2 facility with BSL-3 practices) until the agent is excluded.
- Refer. If the agent cannot be ruled out, stop. Notify the supervisor and the LRN reference laboratory. Do not set up VITEK, Phoenix, MicroScan, API strips, or MALDI-TOF. Automated systems and open-spot MALDI generate aerosols and misidentify these organisms (B. anthracis as B. cereus; Y. pestis as Y. pseudotuberculosis, Shigella, or Acinetobacter; Brucella as former Ochrobactrum or a non-brucellosis Brucella; Francisella as Haemophilus or no identification).
If an automated system unexpectedly names a select-agent species, treat the isolate as not ruled out: move remaining work into a BSC and refer.
Presumptive identification patterns (memorize this table)
| Organism | Gram stain and growth | Colony morphology | Rapid tests that leave it not ruled out |
|---|---|---|---|
| Bacillus anthracis | Large Gram-positive rod; spores rarely seen in clinical material; grows on BAP; no growth on MAC | Nonhemolytic, gray-white ground-glass, medusa-head curling edges, tenacious (stands up like beaten egg white when teased) | Catalase-positive, nonmotile. Beta-hemolytic or motile large GPR → B. anthracis is ruled out |
| Yersinia pestis | Gram-negative rod, often bipolar (safety-pin) on direct smears; slow (pinpoint at 24 h, better at 48 h); grows at 25–28°C and 35–37°C; non-lactose fermenter on MAC | Gray-white, later fried-egg or hammered-copper | Oxidase-negative, urease-negative, nonmotile at 35°C (and at 25°C—other yersiniae are motile at 25°C), catalase-positive, indole-negative |
| Brucella spp. (brucellosis-causing, BBS) | Tiny Gram-negative coccobacilli (may stain Gram-variable); slow 48–72 h; no growth on MAC | Tiny, smooth, raised, translucent, nonhemolytic on BAP/CHOC | Catalase-positive, oxidase-positive, urease-positive. B. suis may split urea in minutes |
| Francisella tularensis | Tiny Gram-negative coccobacillus; cysteine-dependent—grows better on CHOC (or cysteine heart agar) than on BAP; no MAC | Tiny, gray-white, nonhemolytic; may need 48–72 h | Oxidase-negative, catalase weak or negative, beta-lactamase-positive. Better growth on CHOC than BAP is a major clue versus Brucella |
Organism-by-organism rule-out points
B. anthracis. The combination that must not be ignored is a large GPR that is nonhemolytic and nonmotile. B. cereus group organisms are typically beta-hemolytic and motile. Tenacity (the colony stands up as a peak when lifted) and medusa-head spreading are classic but are not required before you trigger referral. Do not vortex, sniff, or India-ink the colony on the open bench. B. cereus biovar anthracis is handled on the same sentinel pathway.
Y. pestis. Think slow enteric that is urease-negative and nonmotile. Y. enterocolitica is urease-positive and motile at 25°C. Fried-egg morphology appears with prolonged incubation; do not wait for it before referring a matching biochemical set. Growth is often better at 25–28°C than at 35°C, which matches the flea-gut temperature of the organism's life cycle.
Brucella spp. Think tiny GNCB from blood that will not grow on MacConkey and is catalase-, oxidase-, and urease-positive. This pattern overlaps Haemophilus, some Acinetobacter, Bordetella, and environmental organisms formerly called Ochrobactrum (now placed in Brucella but not causes of brucellosis). That taxonomic overlap is exactly why ASM shifted sentinel emphasis from "presumptive identification" to rule-out of brucellosis-causing Brucella species (BBS).
Critical current fact—teach both the outline language and the 2024–2025 update. The ASM Brucella sentinel guideline (revised August 2025) states that Brucella was removed as a U.S. select agent. The HHS/USDA final rules that removed the overlap species B. abortus, B. melitensis, and B. suis were published December 17, 2024 and took effect January 16, 2025; ASM's guideline cites the December 17 publication date, so both dates appear in the literature. Deregulation changes biosecurity paperwork—Federal Select Agent Program forms, registered-entity destruction logs, and some Category A shipping triggers—but it does not change biosafety. BBS remains highly infectious, remains a reportable / nationally notifiable disease, and still requires rule-out and referral. Work continues in a BSC with BSL-3 practices. Isolator lysis-centrifugation is no longer specified for specimen processing. Gram-variable staining is now recognized. The ASCP M outline still lists Brucella among BSL-3/select-agent examples, so exam items may use that language; a correct modern answer still keeps full biosafety and referral.
F. tularensis. Think tiny GNCB that prefers CHOC/cysteine, ignores MAC, is oxidase-negative, catalase-weak, and beta-lactamase-positive. Brucella is oxidase- and catalase-positive and does not need extra cysteine. Haemophilus is X/V dependent; satellite growth around a staphylococcus streak is not a Francisella feature. The infectious dose is among the lowest of any bacterium—do not aerosolize it on an automated instrument.
Safety actions that are not optional
- Perform all follow-up of a suspected isolate in a Class II BSC.
- Do not centrifuge without sealed rotors or cups, and open those cups only in the BSC.
- Do not sniff plates.
- Do not prepare slides or MALDI spots on the open bench. If a manufacturer tube-extraction MALDI method is ever used in a reference setting, the extract is filtered (0.2 µm or less); sentinel labs should not be improvising that workflow on a suspect plate.
- Keep a log of who handled the plates before suspicion was raised; the LRN and occupational health need that list if the isolate confirms, because prophylaxis may be indicated.
If rule-out tests exclude the agent, return to the usual identification pathway. If they do not, the laboratory's product is a referral, not a species name on the chart.
A large Gram-positive rod from a blood culture is nonhemolytic, catalase-positive, nonmotile, and forms tenacious ground-glass colonies with medusa-head edges. The correct sentinel action is:
Which rapid-test pattern leaves Francisella tularensis not ruled out?
The ASM Brucella sentinel guideline (August 2025) addresses removal of Brucella as a U.S. select agent, finalized in a rule published December 17, 2024 and effective January 16, 2025. Which statement is correct?