Why Do Toxic Granules Become Conspicuous?

Primary granules are already there. What changes is their stainability.

Introduction

Toxic granulation has been recognized for more than a century. Schofield and colleagues traced early descriptions to Arnold in 1895, Cesaris-Demel in 1909, and Mommsen in 1929. Yet the biological basis of this familiar blood-smear finding remained uncertain well into the modern era.

Medical students learn to associate toxic granulation with bacterial infection. Clinicians report it alongside Döhle bodies and cytoplasmic vacuoles. Hematologists recognize it almost immediately.

But the name creates a misleading picture. The granules do not contain a toxin, nor are they necessarily a newly generated abnormal structure.

The real question is more precise:

Why do primary granules that are normally inconspicuous become so intensely visible?

The conventional answer is that infection accelerates granulopoiesis and releases neutrophils before they have lost their primary granules.

But primary granules are not normally lost.

They are produced early in neutrophil development and remain present in mature circulating neutrophils. What largely fades during normal maturation is their conspicuous azurophilic appearance on a routinely prepared Romanowsky-stained smear.¹ The available cytochemical evidence suggests that this intense staining depends mainly on acidic, apparently sulfated intragranular material that adsorbs Romanowsky dyes, rather than on the granule enzymes themselves.

The problem is therefore not simply one of granule presence.

It is one of granule stainability.

What is toxic granulation?

Schofield and colleagues described toxic granulation as:²

“Prominent staining, by Romanowsky dyes, of cytoplasmic granules in blood neutrophils during an acute inflammatory illness.”

On a Wright- or Wright-Giemsa-stained smear, affected neutrophils contain conspicuous dark blue-purple cytoplasmic granules.

Toxic granulation is one manifestation of broader neutrophil toxic change, which may also include Döhle bodies and cytoplasmic vacuoles.³ These findings frequently coexist, but they have different structural bases. This essay focuses on toxic granulation because its mechanism poses a particularly interesting problem in neutrophil development and cytochemistry.

When are neutrophil granules made?

Neutrophil development proceeds through the following sequence:

Myeloblast → promyelocyte → myelocyte → metamyelocyte → band neutrophil → segmented neutrophil

These stages can be divided into two broad compartments.

Mitotic compartment

• Myeloblast
• Promyelocyte
• Myelocyte

These cells retain the capacity to divide.⁴

Postmitotic compartment

• Metamyelocyte
• Band neutrophil
• Segmented neutrophil

After the myelocyte stage, cell division ceases. Subsequent development consists of terminal maturation rather than additional mitosis.

This distinction matters because changes during granulopoiesis may arise from two different processes:

1. partitioning of cellular contents among daughter cells during mitosis, and
2. maturation of existing cellular structures after mitosis has stopped.

Primary granules are produced early

Primary, or azurophilic, granules form predominantly during the promyelocyte stage. Expression of primary-granule proteins begins around the myeloblast-to-promyelocyte transition and continues through the promyelocyte phase. Their contents include myeloperoxidase, neutrophil elastase, cathepsin G, proteinase 3, azurocidin, defensins, and other antimicrobial proteins.

Specific, or secondary, granules begin forming later, predominantly across the myelocyte–metamyelocyte interval. Gelatinase granules are generated later still, and secretory vesicles appear at the latest stages of development. These boundaries overlap rather than functioning as absolute switches, but the overall sequence is well established.

Modern models therefore explain granule identity primarily through the timing of granule-protein expression. Proteins expressed early enter primary granules; proteins expressed later enter later granule populations.

Most primary-granule biogenesis is confined to the early promyelocytic phase rather than continuing throughout terminal maturation.

The granules are synthesized early and retained throughout neutrophil maturation.

Why the conventional explanation is incomplete

The usual explanation of toxic granulation goes something like this:

During severe infection, the marrow accelerates neutrophil production and releases immature cells before they have lost their primary granules. The retained granules produce the coarse purple stippling seen on the smear.

This explanation contains an important truth. Toxic granulation appears to arise during altered granulopoiesis rather than through the sudden formation of a new abnormal organelle in a mature circulating neutrophil.

But the language of “retention” can be misleading.

Normal mature neutrophils also retain primary granules.

The question is why those granules are ordinarily inconspicuous and why they become so intensely stainable under some conditions.

The staining experiment that reveals the paradox

In 1969, McCall and colleagues studied neutrophils from patients with severe bacterial infections using light microscopy, electron microscopy, cytochemistry, and biochemical assays.⁵

One of their simplest experiments may also have been the most persuasive.

On a routinely prepared Wright-stained smear, mature control neutrophils showed no visible azurophilic granules.

But when normal blood smears were fixed with glutaraldehyde before staining, azurophilic granules became visible and resembled toxic granules.

The same thing happened when the buffering phase of the Wright-staining procedure was extended from four minutes to one hour:

“We could also visualize similar granules in mature control neutrophils by light microscopy, but only after ‘fixation’ in glutaraldehyde or after prolonged exposure to Wright stain.”

Nothing new had appeared inside the cells.

The granules had been present all along.

Changing the preparation and staining conditions had simply made them conspicuous.

That experiment reframed toxic granulation as a problem of tinctorial behavior, meaning how pre-existing granules interact with and retain stain.

What toxic granules are not

On routine Wright-stained blood smears, McCall and colleagues counted an average of approximately 105 visible toxic granules per neutrophil in affected patients and none in controls.

But when they examined the same phenomenon ultrastructurally, the contrast largely disappeared.

The conspicuous azurophilic granules seen by light microscopy corresponded to the large, dense, peroxidase-positive primary granules found in both toxic and control neutrophils.

They were not:

• phagocytosed bacteria or bacterial products
• a newly generated abnormal granule population
• autophagic bodies
• nonspecific cytoplasmic debris.

The investigators also found no statistically significant differences in section-based granule counts:

Method	Measurement	Toxic neutrophils	Controls
Routine light microscopy	Visible toxic granules per neutrophil	105	0
Thick-section light microscopy	Azurophilic granules per section	36	42
Electron microscopy	Total granules per section	190	178
Peroxidase electron cytochemistry	Peroxidase-positive granules per section	98	88

These were not equivalent measurements. The routine-smear count reflected what could be seen after standard Wright staining, whereas the other values were section-based counts obtained by thick-section microscopy or electron microscopy. They were not stereologic estimates of whole-cell granule number.

McCall explicitly acknowledged the sampling limitations and noted that modest quantitative differences could not be excluded.

Still, the contrast was striking:

About 105 versus zero by routine light microscopy, but 36 versus 42 by thick-section microscopy.

The dramatic difference in appearance was not matched by a demonstrated major difference in the number of underlying primary granules.

McCall’s group therefore concluded that toxic granulation represented an abnormal staining reaction of ordinary azurophilic granules. They had identified the relevant organelle, although they remained cautious about whether the broader toxic-neutrophil phenotype reflected immaturity, stimulation, degeneration, or some combination of these processes.

Direct bacterial exposure did not reproduce the finding

Schofield and colleagues later incubated normal neutrophils with opsonized Staphylococcus aureus at 10 bacteria per cell and, separately, with Escherichia coli endotoxin for as long as 180 minutes.

Neither exposure produced toxic granulation.

This does not prove that no circulating or systemic factor can ever influence granule appearance. A short in-vitro experiment cannot reproduce the full marrow, cytokine, stromal, and vascular environment of severe infection.

Together with the maturation data, however, the result favors establishment during altered marrow development rather than direct conversion of a mature normal neutrophil by brief exposure to bacteria or endotoxin.

A generalized increase in permeability is also unlikely

One possible explanation would be that toxic neutrophils stain more intensely because dyes or substrates enter the cells or granules more readily.

Schofield examined the permeability of intact neutrophils to the exogenous chloroacetate-esterase substrate naphthol AS-D chloroacetate.

Toxic neutrophils showed lower, not higher, assay permeability than normal neutrophils:

Measurement	Toxic neutrophils	Normal neutrophils	Significance
Chloroacetate-substrate permeability	6.5	14.1	P < 0.02

This assay did not directly measure Romanowsky-dye entry into individual azurophilic granules. It does, however, argue against a generalized acquired increase in cellular or granule permeability as the explanation for toxic granulation.

The quantitative cytochemistry study

Fourteen years after McCall’s study, Schofield and colleagues approached the problem with quantitative cytochemistry.

Using scanning-integrating microdensitometry, they examined:

• 1,000 toxic-granulation neutrophils from 20 patients with bacterial infection
• 1,250 normal blood neutrophils
• myeloid precursors from 10 normal marrows.

They measured several constituents associated with primary granules:

• myeloperoxidase
• β-glucuronidase
• acid phosphatase
• chloroacetate esterase
• Alcian blue staining for acid mucosubstance

They also measured lactoferrin, a marker of specific granules.

This allowed them to ask two related but different questions:

1. What happens to primary-granule constituents during normal maturation?
2. What distinguishes toxic from normal mature neutrophils?

What normally changes during maturation?

Across normal marrow maturation, the measured primary-granule enzymes declined, while lactoferrin increased.

Measurement, mean OD units	Promyelocyte	Myelocyte	Metamyelocyte	Mature polymorph
Myeloperoxidase	102.4	72.2	48.6	44.4
β-glucuronidase	63.8	44.5	33.9	29.6
Acid phosphatase	49.6	34.9	26.9	22.8
Chloroacetate esterase	108.6	76.7	58.9	51.6
Alcian blue	9.1	8.1	1.8	0.4
Lactoferrin	32.9	31.7	39.0	50.9

These were cross-sectional cytochemical measurements in morphologically classified individual cells, not longitudinal measurements of the same cells over time. The optical-density values reflect total reaction product per cell. They do not distinguish among changes in granule number, the amount of a constituent within each granule, or its accessibility to the cytochemical stain.

Some reduction in primary-granule constituents may reflect events in the mitotic compartment. Because primary granules are produced early, later divisions may partition the existing granule endowment among daughter cells. Meanwhile, later granule populations accumulate and may dilute the relative prominence of the primary-granule compartment.

Schofield and colleagues therefore proposed several possible explanations for the decline in acid mucosubstance staining:

• masking of the mucosubstance
• relocation from the granule cortex to its core
• mitosis together with dilution as specific granules accumulated
• some combination of these processes.

Their study did not directly determine which mechanism predominated.

Something continues to change after mitosis stops

The postmitotic interval provides an important clue.

From the metamyelocyte to the mature polymorph stage, Alcian blue staining fell from 1.8 to 0.4, a significant decline with P < 0.02.

Across that same entirely postmitotic interval, none of the four measured primary-granule enzymes changed significantly.

Because metamyelocytes, bands, and segmented neutrophils no longer divide, mitotic partitioning cannot explain this late decline. This observation also has an important implication. Simply releasing these postmitotic cells earlier into the circulation cannot, by itself, explain why their primary granules remain conspicuous. Whatever produces toxic granulation must reflect altered development within the marrow rather than the timing of marrow release alone.

The findings therefore indicate that the loss of acid mucosubstance staining continues after mitosis has ceased. The cytochemical data do not reveal what drives this change. It may represent execution of a maturation program established earlier in granulopoiesis, modification of that program by the marrow microenvironment, including cytokines and stromal signals, or some combination of the two.

Nor do the data identify the actual intragranular change responsible for the loss of staining. Possibilities include:

• removal of the acidic material
• chemical modification
• masking
• relocation within the granule
• altered accessibility to stain

What the study establishes is the timing of the observable change, not the molecular mechanism responsible for it.

The cleanest comparison: toxic versus normal mature neutrophils

The most revealing comparison was between toxic and normal mature blood neutrophils.

Granule compartment	Measurement, mean OD units	Toxic neutrophils	Normal neutrophils	Significance
Primary granules	Myeloperoxidase	44.3	44.3	NS
Primary granules	β-glucuronidase	30.7	30.0	NS
Primary granules	Acid phosphatase	26.0	18.4	P < 0.01
Primary granules	Alcian blue staining for acid mucosubstance	5.0	0.1	P < 0.01
Secondary granules	Lactoferrin	39.1	48.9	NS

Myeloperoxidase and β-glucuronidase, both enzymatic constituents of primary granules, were essentially identical in toxic and normal neutrophils.

These were cell-level cytochemical measurements. They did not directly measure the number of primary granules or the amount of each constituent within an individual granule.

Nevertheless, the results argue against a major increase in the number of primary granules as the principal explanation for toxic granulation. If toxic neutrophils contained substantially more primary granules because those granules had undergone less mitotic dilution, one would expect at least some corresponding increase in broadly distributed primary-granule constituents such as myeloperoxidase or β-glucuronidase. Neither was observed.

The clearest measured abnormality was the marked persistence of acidic, apparently sulfated intragranular material detected by Alcian blue staining.

What is acid mucosubstance?

Schofield and colleagues characterized the Alcian blue-reactive material using two complementary histochemical stains. Alcian blue demonstrated acid mucosubstance, whereas high-iron diamine supported the presence of sulfated mucosubstance.

High-iron diamine staining was positive in an average of 92.3% of polymorphs from nine patients with toxic granulation and absent from the normal control polymorphs studied.

The authors described the material as acidic and presumably sulfated. Prior work suggested that such mucosubstance could help provide stable storage conditions by complexing with lysosomal hydrolases.

As normal neutrophils matured, Alcian blue reactivity and conspicuous blue-purple Romanowsky staining declined in parallel.

In toxic neutrophils, persistence or greater accessibility of this acidic, apparently sulfated intragranular material is the best-supported explanation for the intense Romanowsky staining.

The study did not prove whether the material was physically removed, masked, relocated, chemically modified, or processed in some other way.

Nor did it directly measure the amount of mucosubstance within each individual granule. The cytochemical values were measurements at the cellular level.

The cytochemical signal tracks what observers saw

Schofield did not merely show that Alcian blue staining was elevated in toxic neutrophils.

The Alcian blue score correlated with the morphologic toxic-granulation score in all three groups examined:

• infected patients: r = 0.64, P < 0.02
• uninfected patients with leukocytosis: r = 0.64, P < 0.01
• leukemic patients: r = 0.84, P < 0.01.

This directly links the cytochemical variable to the visual finding graded under the microscope.

The more acidic mucosubstance the neutrophils retained, the more toxic granulation observers saw.

Why do toxic granules look “coarse”?

Toxic granules are commonly described as coarse, and some morphology resources describe them as appearing larger.

McCall’s electron-microscopic analysis, however, found no significant difference when granules were divided into large, medium, and small categories.

Thus, the apparently coarse appearance seen by light microscopy may not represent a true increase in granule dimensions.

A plausible explanation is optical. At the resolution limit of light microscopy, a granule that binds substantially more dye may appear denser, broader, and more coarsely defined.

This remains an inference rather than a directly proven mechanism, but it reconciles the apparent enlargement by light microscopy with the absence of a demonstrated size difference by electron microscopy.

They may look coarser without actually being larger.

A possible modern molecular correlate

Schofield described the relevant material as “acid, presumably sulphated, mucosubstance.”

Modern neutrophil biology offers a possible molecular framework.

Serglycin is a sulfated proteoglycan involved in the retention and protection of proteins within neutrophil granules. Together with cathepsin C, it helps preserve granule proteins and maintain their intragranular localization.⁶

This makes a serglycin-containing proteoglycan matrix one plausible molecular correlate of the material detected histochemically in 1983.

But this remains a hypothesis.

None of the studies reviewed here demonstrates that toxic granulation results from altered:

• serglycin abundance
• glycosaminoglycan sulfation
• serglycin processing
• intragranular localization
• postmitotic remodeling.

The modern question is therefore specific and experimentally testable:

What happens to sulfated proteoglycan-associated material during normal postmitotic primary-granule maturation, and why is that process altered during inflammatory granulopoiesis?

Toxic granulation and left shift

Toxic granulation commonly accompanies a left shift, but the two findings are not synonymous.

A left shift refers to the increased appearance of earlier granulocytic forms—particularly bands and metamyelocytes—in the peripheral blood.

Toxic granulation refers to altered staining of primary granules within neutrophils.

The two frequently coexist because both arise during inflammatory granulopoiesis, but they describe different aspects of neutrophil development.

Toxic granulation is not specific to infection

Toxic granulation is strongly associated with bacterial infection, but it is not entirely specific.

Schofield also studied:

• 18 uninfected patients with neutrophilic leukocytosis caused by postoperative states, nonhematologic malignancy, cardiovascular disease, prednisolone use, and other conditions
• seven patients with acute nonlymphoid leukemia in remission
• three patients with treated, steady-state chronic granulocytic leukemia.

The uninfected leukocytosis group showed weak but significantly increased Alcian blue staining and mild toxic granulation. The leukemic group also showed increased Alcian blue reactivity and modest toxic granulation, although the changes were considerably less pronounced than in patients with bacterial infection.

These findings support the interpretation that toxic granulation reflects altered granulopoiesis or abnormal granule maturation rather than direct bacterial modification of circulating cells.

The classic studies were highly selected for patients who already had prominent toxic granulation. They do not establish its sensitivity, specificity, or prognostic value for modern clinical practice.

Where do Döhle bodies and vacuoles fit?

Toxic granulation is one component of neutrophil toxic change.

McCall demonstrated that Döhle bodies correspond to lamellar aggregates of rough endoplasmic reticulum. By correlating Wright-stained thick sections with adjacent electron-microscopic sections, the investigators identified multiple rows of rough endoplasmic reticulum within the pale blue inclusions.

Cytoplasmic vacuoles represented another distinct alteration. In McCall’s experiments, vacuolation increased during incubation, was associated with altered lysosomal activity and degranulation, could be promoted by endotoxin or phagocytosis, and occasionally contained autophagic material.

These findings often coexist during severe inflammation, but they do not share a single structural explanation.

Toxic granulation is specifically a question of why ordinary primary granules become intensely stainable.

A revised model

The evidence supports the following model:

1. Primary granules are produced predominantly during the promyelocyte stage.
2. Myeloblasts, promyelocytes, and myelocytes comprise the mitotic compartment. During these stages, the existing primary-granule endowment may be partitioned among daughter cells.
3. Later granule populations accumulate as differentiation proceeds. This may reduce the relative prominence of primary granules within the total granule population.
4. After the myelocyte stage, cells enter the postmitotic compartment. Metamyelocytes, bands, and segmented neutrophils mature without further division.
5. During normal postmitotic maturation, acid mucosubstance associated with primary granules becomes progressively less detectable. The mechanism may involve masking, relocation, chemical modification, or some other change in accessibility.
6. Normal mature neutrophils therefore retain primary granules that are usually inconspicuous on routine Romanowsky staining.
7. In toxic neutrophils, this maturational loss of stainability is altered. Acidic, apparently sulfated intragranular material persists or remains accessible, allowing the pre-existing primary granules to adsorb Romanowsky dye more intensely.
8. A major increase in primary-granule number or size has not been demonstrated. Modest quantitative differences cannot be excluded because the available ultrastructural counts were section-based and limited by sampling.
9. The most clearly demonstrated change is tinctorial. The granules become far more visible than their ultrastructural number or measured enzyme content would predict.

Toxic granulation is therefore best understood as:

Increased visibility of pre-existing azurophilic granules caused by abnormal persistence or accessibility of acidic, apparently sulfated intragranular material.

Primary granules: what changes and what does not

Feature	Steady-state granulopoiesis	Infection-associated granulopoiesis
Primary granule size	No major change demonstrated	No major change demonstrated
Primary granule number	No major change demonstrated*	No major change demonstrated*
Primary granule stainability	Progressively decreases	Persists

*Modest differences cannot be excluded because available ultrastructural counts were section-based and limited by sampling.

The remaining question

The classic studies answered several important questions.

Toxic granules correspond to ordinary primary granules.

They are not phagocytosed bacteria.

They are not a newly formed abnormal organelle.

A major increase in their number or size has not been demonstrated.

Myeloperoxidase and β-glucuronidase content are essentially the same in toxic and normal mature neutrophils.

What differs most clearly is the persistence of acidic, apparently sulfated intragranular material and the resulting capacity to bind Romanowsky stain.

But the decisive molecular step remains unknown.

What normally causes this material to become masked, relocated, modified, or otherwise inaccessible during postmitotic maturation?

Does the process involve serglycin or another sulfated proteoglycan?

What changes during infection-associated granulopoiesis to preserve the stain-reactive state?

The granules themselves are not the mystery.

The unresolved question is how a primary granule that is ordinarily almost invisible becomes toxic granulation.

Suggested key references

1. McCall CE, Katayama I, Cotran RS, Finland M. Lysosomal and ultrastructural changes in human “toxic” neutrophils during bacterial infection. J Exp Med. 1969;129(2):267–293. The paper established that toxic granules correspond to ordinary azurophilic granules, documented the altered Wright-staining behavior, and found no clearly demonstrated major difference in section-based granule counts.
2. Schofield KP, Stone PCW, Beddall AC, Stuart J. Quantitative cytochemistry of the toxic granulation blood neutrophil. Br J Haematol. 1983;53:15–22. This is the central cytochemical study linking toxic granulation to persistence of acidic, apparently sulfated mucosubstance in azurophilic granules.
3. Yin C, Heit B. Armed for destruction: formation, function and trafficking of neutrophil granules. Cell Tissue Res. 2018;371(3):455–471. Use the uploaded modern neutrophil-granule review as the reference for the temporal model of granule biogenesis and the discussion of serglycin and cathepsin C.